Skip to main content

Full text of "Proceedings of the Linnean Society of New South Wales"

See other formats


THE 



PROCEEDINGS 



LiriNEnri Society 



New South Wales 



FOR THE YEAR 

1959 

VOL. LXXXIV. 



WITH TWENTY-THREE PLATES. 

287 Text-figures. 



Sydney: 

printed and published for the society by 

AUSTRALASIAN MEDICAL PUBLISHING CO. LTD. 

Seamer Street, Glebe, Sydney, 

and 

SOLD BY THE SOCIETY. 

19(W. 



CONTENTS. 



CONTENTS OF PROCEEDINGS, 1959 



PART 1 (No. 389). 
(Issued 29th Jiihj. 1959.) 

Pages 

Presidential Address, delivered at the Eighty-fourth Annual General Meeting, 
25th March. 1959, by Dr. S. Smith-White: 

Summary of Year's Activities 1-4 

Pollen Development Patterns in the Epacridaceae. A Problem in 

Cytoplasm-Nucleus Interaction. (Plates i-ii: twelve Text-figures.) S- 35 

Elections 4 

Balance Sheets for the Year ending 2Sth February, 1959 .5-7 

Physiological Specialization of Melampsora lini (Pers.) Lev. in Australia. By 

H. B. Kerr. (Plates iii-iv; three Te.xt-flgures.) 36- 63 

The Families of Cycads and the Zamiaceae of Australia. By L. A. S. Johnson. 

(Four Text-figures.) .... .. 64-117 

Root Parasitism in Atkinso7iia ligustrina (A. Cunn. ex F. Muell.) F. Muell. 

By Barbara P. Menzies and H. S. McKee. (Fourteen Text-figures.) .. llS-127 

Descriptions of Two New Species of Citris and One New Species of Stigmodera 

(Buprestidae). By C. M. Deuquet. (Three Text-figures.) 12S-130 

Notes on Australian Mosquitoes (Diptera. Culicidae). IV. Aedes alboanjiulaius 

Complex in Victoria. By N. V. Dobrotworsky. (Four Text-figures.) .. 131-145 



CONTENTS. 



PART 2 (No. 390). 
(Issued 28th October, 1959.) 



Pages 

Observations on the Ecology of the Phasmatid CtenoviorpJiodes tessulata 

(Gray). By P. Hadlington and F. Hoschke. (Plates v-vi; two Te.xt-figures.) 146-159 

An Annotated Catalogue of Described Australian Tabaninae (Diptera, 

Tabanidae). By I. M. Mackerras. (Nine Text-figures.) 160-185 

Observations on Some Australian Forest Insects. 4. Xyleborus truncatus 
Erichson, 1842 (Colooptera: Scolytidae) associated with Dying Eucalyptus 
saligna Smith (Sydney Blue-gum). By K. M. Moore. (Plate vii; three 
Text-figures.) 186-193 

The Genus Conostylis R.Br. I. Leaf Anatomy. By J. W. Green. (Twenty-five 

Text-figures.) 194-206 

Somatic Hybridization between Puccinia graminis var. tritici and Puccinia 

gram.inis var. secalis. By I. A. Watson and N. H. Luig 207-208 

Diptera of Katoomba. Part 3. Stratiomyiidae and Tachinidae. By G. H. Hardy 209-217 

Some East Australian Sea-grass Communities. By E. J. Ferguson Wood. (Two 

Text-figures.) 218-226 

The Effects of Inorganic Salts on Dividing Cells. By Mary M. Hindmarsh. 

(Plate viii; two Text-figures.) 227-231 

The Reproduction and Early Life History of the Gastropod Cymatilesta 
spengleri (Perry) (Fam. Cymatidae). By D. T. Anderson. {Communi- 
cated by Miss I. Bennett.) (Plate ix; seventeen Text-figures.) 232-237 

A New Mite Parasite {Harpyrhynclius) from the Roselle Parakeet (Trombidi- 
formes, Acari). By R. F. Lawrence. {Communicated by Br. A. R. 
Woodhill.) (Two Text-figures.) 238-241 

Floral Structure and Anatomy in the Family Goodeniaceae Dumort. By R. C. 

Carolin. (Forty-five Text-figures.) 242-255 

Iron Deficiency in Eucalyptus dives Schauer. By W. D. Andrew and D. J. 

David. {Communicated by Professor L. D. Pryor.) 256-258 

Pollen Tetrad Segregation in Astroloma pinifoliiim and in Acrotriche fasciculi- 
flora. By S. Smith-White. (Four Text-figures.) 259-270 

Pollen Tetrad Patterns in Leschenaultia. By P. G. Martin and W. J. Peacock. 

(Plate X.) 271-277 

Seed Coat Anatomy and Taxonomy in Eucalyptus. II. By E. Gauba and L. D. 

Pryor. (Plates xi-xiii; Text-figures 20-39.) 278-291 

Experimental Crossing of Aedes {Stegomyia) aegypti Linnaeus and Aedes 
{Stegomyia) albopictus Skuse (Diptera, Culicidao). By A. R. Woodhill. 
(Plate xiv.) 292-294 



CONTENTS. 



PART 3 (No. 391). 
(Issued J2nd February, I'.HUI.) 

Pages 
Ranuncvhis lappaceus and Allied Species of the Australian Mainland. I. 

Taxonomy. By Barbara G. Briggs. (Plate xv; ninety-one Text-figures.) 295-324 

Dindoxus erytJirurus (White) (Coleoptera-Buprestidae), Attack of Fire-damaged 

Callitris spp. By P. Hadlington and M. J. Gardner. (Plate xvi.) .. .. 32.5-331 

Woody Gall, a Suspected Virus Disease of Rough Lemon and Other Citrus 

Varieties. By Lilian R. Fraser. (Plate xvii.) 332-334 

Light as a Control in the Germination and Development of Several Mistletoe 

Species. By J. F. Rlgby 335-337 

A Revision of Nitella cristata Braun (Characeae) and its Allies. Part I. 
Experimental Taxonomy. By T. C. Chambers and Mary B. Williams. (Plate 
xviii; three Text-figures.) 338-345 

A Revision of NiteUa cristata Braun (Characeae) and its Allies. Part II. 

Taxonomy. By Mary B. Williams. (Six Text-figures.) 346-355 

Some Silurian Lamellibranchs from New South Wales. By Kathleen Sherrard. 

(Plate xix.) 356-372 

A New Species of Hilara (Diptera, Empididae) from New South Wales. By 
Kenneth G. V. Smith. {Communicated by Mr. C. E. Chadioick.) (Four 
Text-figures.) 373-374 

The Anatomy of the Barks of Five Species of Callitris Vent. By R. K. Bamber. 

(Plates xx-xxi.) 375-381 

Some Species of Culex {Loplioceraoviyia) from New Guinea and Adjacent 
Islands, with Descriptions of Four New Species and Notes on the Male 
of Culex fraudatrix Theobald (Diptera, Culicidae). By Donald H. Colless. 
(Three Text-figures.) 382-390 

Notes on Australian Thynninae. IV. The Morphology of Thynnoides ruflt^wrax 
Turner, with Notes on the Prepupal Larva and the Cocoon. By B. B. 
Given. {Communicated by Dr. A. J. 'Nicholson.) (Twenty-three Text- 
figures.) .. 391-399 

Notes on Australian Thynninae. V. Aeolotliynnus m.ultiguttatus Ashmead and 
Isioaroides koebelei Ashmead. By B. B. Given. {Communicated by Dr. 
A. J. Nicliolson.) (Eight Text-figures.) 400-402 

William Noel Benson, 1885-1957. (Memorial Series, No. 16.) (With Portrait, 

Plate xxii.) 403-409 

Douglas Mawson, 1882-1958. (Memorial Series, No. 17.) (With Portrait, Plate 

xxiii.) 410-414 

Abstract of Proceedings jjs"" ~ '"" 415-421 

List of Members . . /^'^ . . . , A . . . 422-427 

^- ^ ? ^ A 

List of Plates M . .^:P^ . f\ 400 

1%. % ' 
List of New Family, Species and Subspecies . . \%i C j 428 

Index .^ ' ... . . 429-432 



S.cS/^6.^, 



{Issued 29th July, 1959.) 



ac=s> 



<=3n 



I 



Vol. LXXXIV. 
Part 1. 



No. 389. 



THE 



PROCEEDINGS 



LiriNEnn Society 



or 



New South Wales 



FOR THE YEAR 



1 959 



Part 1 (pp. 1-1 Jf5). 

COWTAINING THE PROCEEDINGS OP THE ANNUAL, MEETING ANO 
PAPERS READ IN MARCH-APRIL,. 

With four plates. 
[Plates i-iv.] 




Sydney: 

printed and published foe the society by 

AUSTRALASIAN' MEDICAL PUBLISHING CO. LTD.. 

Seamer Street, Glebe, Sydney, 

and 

SOLD BY THE SOCIETY, 

Science House, 157 Gloucester Street, Sydney. 



Registered at the General Post Office, Sydney, for tranfsmisslon 
by post as a periodical. 



Agent in Europe: 
David Nutt, 41 Colebrooke Row, London, N.l. 



The Linnean Society of New South Wales 



LIST OF OFFICERS AND COUNCIL, 1959-60. 



President: 

T. G. Vallance, B.Sc, Ph.D. 

Vice-Presidents ; 

S. J. Copland, M.Sc. F. V. Mercer, B.Sc, Ph.D. 

Lilian Fraser, D.Sc. S. Smith-White, D.Sc.Agr. 

Hon. Treasarer: 

A. B. Walkom, D.Sc. 

Hon. Secretaries: 

W. R. Browne, D.Sc; A. B. Walkom, D.Sc. 

Coonetl: 

R. H. Anderson, B.ScAgr. S. Smith-White, D.Sc.Agr. 

W. R. Browne, D.Sc. E. Le G. Troughton, C.M.Z.S., F.R.Z.S. 

A. N. Colefax, B.Sc. T. G. Vallance, B.Sc, Ph.D. 

S. J. Copland, M.Sc. J. M. Vincent, D.Sc.Agr., Dip.Bact. 

Professor R. L. Crocker, D.Sc. A. B. Walkom, D.Sc. 

J. W. Evans, M.A., D.Sc, Sc.D. H. S. H. Wardlaw, D.Sc, F.R.A.C.I. 

Lilian Fraser, D.Sc. Professor W. L. Waterhouse, C.M.G., 
F. V. Mercer, B.Sc, Ph.D. M.C., D.ScAgr., D.I.C. 

I. V. Newman, M.Sc, Ph.D. A. R. Woodhill, D.Sc.Agr. 
Professor B. J. F. Ralph, B.Sc, Ph.D., 
A.A.C.I. 

Aaditor; 

S. J. Rayment, F.C.A. (Aust.). 



NOTICE 

Sets of the Pkoceedings of the Linnean Society of New South Wales (except 
First Series, Volumes I- VI, 1875-1881), and separate Parts, may be purchased from the 
Society, Science House, 157 Gloucester Street, Sydney. Separate Parts may be obtained 
also from David Nutt, 41 Colebrooke Row, London, N.l. The stock of First Series, 
Volumes I-VI, is strictly limited and some Parts of these Volumes are out of print. 

From 1st March, 1954, the price of each complete volume will be £3 10s., plus postage. 
(All previous price lists are cancelled.) 

Index to Volumes I-L of the Proceedings [Issued 15th February, 1929]. Pp. 108. 
Price 5s. 

The Macleay Memoeial Volume [Issued 13th October, 1893]. Royal 4to, ii to 308 
pages, with portrait, and forty-two plates. Price £1 Is. 

Descriptive Catalogue of Australian Fishes. By William Macleay, F.L.S. [1881]. 
A tew copies only. Two volumes and supplement. Price £2 2s. 

The Tsansaotions of the Entomological Society of New South Wales, 2 vols., 8vo. 
LVol. I (complete in five parts, 1863-66), price 70s. net. Parts 2-5 10s. each; Vol, II 
(complete in five parts, 1869-73), price 30s. net, or single Parts 7s. 6d. each.] 



SUBSCRIPTION: £S IDs. per annnm; postage 2s. 



ANNUAL GENERAL MEETING. 
25th Maech, 1959. 




The Eighty-Fourth Annual General Meeting was held in the Society's Rooms, 
Science House, 157 Gloucester Street, Sydney, on Wednesday, 25th March, 1959. 

Dr. S. Smith-White, President, occupied the chair. 

The minutes of the Eighty-Third Annual General Meeting, 26th March, 1958, were 
read and confirmed. 

Presidential Address. 

Presidency of the Linnean Society of New South Wales confers both an honour 
and an obligation. I should like to express my thanks to the Society for the former. 
I feel that there is an obligation on the President, in presenting the address, to 
maintain the standards of past years. I hope I will not fail the Society in this respect. 

I should like to thank Dr. W. R. Browne and Dr. A. B. Walkom for their continued 
services as Honorary Secretary and Honorary Treasurer-Editor. They have served us 
for so long and have served us so well that we must be careful not to take their 
services for granted. I should also like to express appreciation to Miss G. L. Allpress, 
our Assistant Secretary, for her continued loyalty; the efficient management of the 
daily affairs of our Society and the maintenance of our Library reflect our indebtedness 
to her. 

The Society's Proceedings for 195.S, Volume S3, Parts 1 and 2, were published in 
1958, and Part 3 in March, 1959. Volume 83 consists of 387 pages, 6 plates and 392 
text-figures. A grant of £125 was made by the Commonwealth Publications Fund 
Committee towards the cost of publication of "Catalogue of Australian Mammals and 
their Recorded Internal Parasites", by Dr. M. Josephine Mackerras, and payment for a 
two-colour plate to illustrate a paper by E. Gauba and L. D. Pryor, entitled "Seed 
Coat Anatomy and Taxonomy in Eucalyptus. I", was made by the Department of the 
Interior, Canberra, A.C.T. Printing of illustrations by collotype was considered by 
Council and it was resolved that appropriate action be taken to let prospective authors 
know of the availability of collotype illustrations for their papers, under certain 
conditions, including one that the author bear half the cost above that of a corresponding 
half-tone block. A minimum number of eight plates was another proviso, as for any 
number less than eight the cost would be prohibitive. 

During the year twelve new members were added to the list, two members died, 
nine members and one associate member resigned, and two were removed from the 
list under Rule vit. The numerical strength of the Society at 28th February, 1959, 
was: Ordinary Members, 223; Life Members, 32; Corresponding Members, 2; total, 259. 

The resignation from Council of Mr. A. J. Bearup was regretfully accepted by 
Council at its February, 1959, meeting and Dr. I. V. Newman was elected in his place. 
Mr. Bearup resigned on account of ill health. We greatly appreciate his services to 
and interest in the Society and offer him our best wishes for his speedy recovery. 

New holland blinds for our suite of offices and meeting room, a steel filing cabinet 
and twenty-five new, comfortable, modern seats have been purchased during the year. 

On 19th June, 1958, the first Sir William Macleay Memorial Lecture was delivered 
by Professor A. A. Abbie, of Adelaide, to a large audience in the hall in Science House. 
The title of the lecture was "Timing in Human Evolution" and the preliminary part 
was devoted to interesting facts in the lives of the Macleays, and Sir William Macleay 
in particular (see Proceedings, S3, 1958: 197-213). 

Proceedings of the Linnean Society of New South Wales, 1959, Vol. Ixxxiv, Part 1. 

A 



2 PRESIDENTIAL ADDRESS. 

Lecturettes were given at the following meetings: April, National Parks in New 
South Wales, by Mr. Allen A. Strom; September, The Natural History of Australian 
Echinoderms, by Miss Elizabeth. C. Pope; and October, Whales and Whaling, by Mr. 
W. H. Dawbin. No ordinary monthly meetings were held in June or August, 1958. 
A symposium was held in July, 1958, on "Origin and Distribution of Australian Fauna 
and Flora", in which a leading part was taken by Professor R. L. Crocker, Dr. 
J. W. Evans and Mr. Gordon Packham. We express our thanks and appreciation to 
all lecturers for their contributions to the programmes of the meetings. Members 
also showed keen interest in bringing notes and exhibits and eniering into the 
interesting discussions which often took place. 

The relatively small attendances at ordinary monthly meetings have been causing 
the Council some concern, and in an endeavour to meet the convenience of members 
the time of meeting was altered to 6 p.m., without, howeA'er, any marked improvement 
resulting. The matter of increasing attendances is still under consideration and 
suggestions by members would be welcomed. 

Library accessions from scientific institutions and societies on the exchange list 
amounted to 1,727 compared with 1,858 in the previous year. Requests for library 
loans from members and institutions were as numerous as in the past years. The 
following donations to the library were made: (1) Sir Gavin de Beer, two books, 
"The Origin of Species", by Charles Darwin (a reprint of the sixth edition with a 
preface by Sir Gavin de Beer) (1958), and "Evolution by Natural Selection", by 
Charles Darwin and Alfred Russell Wallace (with a foreword by Sir Gavin de Beer) 
(1958); and (2) Captain R. Lloyd Jones, three books by Alvin Boyd Kuhn, Ph.D., 
"The Lost Light" (1940), "Who is this King of Glory?" (1944) and "Shadow of the 
Third Century" (1949). Council decided to purchase "Scientific Serials in Australian 
Libraries" and amendment sheets as published from year to year. Exchange relations 
were entered into with a number of institutions. The Proceedings were offered to 
Museo Civico di Storia Naturale di Venezia, Venice, Italy; Universidad de Buenos 
Aires, Facultad de Ciencias Exactas y Naturales de Buenos Aires, Argentina, and 
North Queensland Naturalists' Club, Cairns, North Queensland (in addition to Abstract 
of Proceedings already sent). Reprints were offered in exchange to: Service de 
Parasitologie Vegetale, Institut Pasteur, Paris, France (Zoological and Entomological); 
Institute of Entomology, Academ.ia Sinica, Peking, China (Entomological); Facultas 
Rerum Naturalium Universitatis Comenianae, Bratislava, Czechoslovakia (Botanical) ; 
National Geological Library, Peking, China (Geological) ; Universite de Dakar, Paculte 
des Sciences, Dakar, French West Africa (Marine Zoology) ; Museo de Historia Natural 
"Javia Prado", Universidad Nacional Mayor de San Marcos, Lima, Peru (Geological); 
and Mammals Research Institute, Polish Academy of Sciences, Bialowieza, Poland 
(Zoological and Entomological). A resumption of the exchange of publications with 
Naturwissenschaftlicher Verein, Hamburg, Germany, took place during the year. 

The total net return from the Society's one-third ownership of Science House for 
the year was £1,303 10s. 9d. A second Neon sign advertisement has been erected on 
Science House. 

The Society contributed the sum of £5 5s. stg. towards the commemoration of the 
Bicentenary of the foundation of the Royal Botanic Gardens, Kew, England. 

Linnean Macleay Fellowships. 

In November, 1957, the Council reappointed Miss Nola J. Hannon and Mrs. 
Mary B. Williams to Fellowships in Botany for 1958. 

Miss Hannon resigned her Fellowship on 11th April, 1958, to take up an appoint- 
ment as a lecturer in the N.S.W. University of Technology (now University of New 
South Wales). She has since obtained her Doctorate of Philosophy. 

During 1958 Mrs. Williams continued a study of the freshwater algal family 
Characeae. This work was drastically interrupted when the Department of Botany 
of the University of New England was destroyed by fire in February, 1958. The whole 



PRESIDENTIAL ADDRESS. 5 

of her personal collection of herbarium material and slides was lost. Part of the 
collection of Characeae of the N.S.W. National Herbarium was recovered in severely- 
damaged condition and that of the Melbourne National Herbarium in moderately 
damaged condition. Fortunately, a large proportion of the data already accumulated 
from this material was saved. The loss of the material itself, however, has been a 
severe handicap to making a complete .study of the group. During the remainder of 
the year an attempt was made to sort through the data on hand, but many gaps were 
found and the work has not been completed. Meanwhile, in collaboration with 
Mr. T. C. Chambers, certain races of Nitella cristata Al.Br. were grown in various 
environments differing in factors of light, temperature, and photoperiod. The results 
of this experiment will be incorporated in a joint paper to be presented later. Mrs. 
Williams did not seek reappointment to a Fellowship for 1959. 

In November, 1958, Miss Alison McCusker, M.Sc, was appointed to a Fellowship 
in Botany for 1959. Miss McCusker proposes to undertake cytological investigations 
in the family Epacridaceae, including (1) the sexual system of Leucopogon 
melaleueoides (A. Cunn. ex DC), (2) a comparison of pollen developments and 
breeding systems in populations of Astroloma pinifolium (Benth.) in eastern New 
South Wales and western Victoria, and (3) the production of two types of pollen in 
Acrotriche divaricata (R.Br.). We extend to Miss McCusker our best wishes for a 
successful year's research work. 

Linnean Macleay Lectureship in MicroMology. 

Dr. Y. T. Tchan, Linnean Macleay Lecturer in Microbiology, University of Sydney, 
has furnished a report of his activities for the year ending 31st December, 1958, as 
follows: Teaching and administrative duties occupied a large proportion of the year 
but time was found for the following research: (1) Soil fertility test. This work 
has been continued. Experiments showed that the algal assay technique can be used 
for the estimation of nitrogen, phosphorus, sulphur and magnesium. In addition the' 
possibility of estimating certain trace elements has been under investigation. Many 
difficulties still have to be overcome but it is likely that the method may prove 
suitable for iron, manganese and molybdenum. A paper has been accepted for 
publication (Plant and Soil, 1959). (2) The study of nitrogen fixation by free living 
micro-organisms has made some progress. A paper has been published on the subject 
of cytochromes of Azotobacter (Moss and Tchan, Peoc. Linn. Soc. N.S.W. , Ixxxiii, 1958, 
Part 2, p. 161). (3) With the N-fixing algae, work is in progress to obtain a general 
technique for isolation of bacteria-free algal cultures using antibiotics. 

Obituaries. 
It is recorded -ijvith regret that the following members died during the year: 

Sir Douglas Mawson, D.Sc, B.E., F.R.S., an Honorary Life Member of the Society, 
who had been a member since 1905, died in Adelaide, South Australia, on 14th October, 
1958. He was born in 1882 at Frizinghall, near Bradford, England, came to Australia 
as a boy, and was educated at Fort Street School, Sydney. He graduated B.E. (1902) 
and B.Sc. (1905) at the University of Sydney and D.Sc. (1909) at the University of 
Adelaide. He contributed only two papers to the Society's Proceedings (in 1905 and 
1906) but always retained a keen interest in the Society in spite of his residence in 
South Australia and his many sojourns in Antarctica. For fuller biographical accounts 
of Sir Douglas Mawson see Aust. J. Sci., v. 21, No. 5, p. 134 (December, 1958) and 
Nature, v. 182, No. 4645, p. 1273 (8th November, 1958). 

Mr. Donald Ford Ross died in Sydney on 4th July, 1958. He had been a member 
of the Society since 1945. 

Dr. Walter George Woolnough, who died on 2Sth September, 1958, at the age of 
82, was a notable figure in the geological life of Australia. He had filled with 
distinction important academic positions in Adelaide, Sydney and Perth, and was the 
first Geological Adviser to the Commonwealth Government. He joined this Society in 



4 PRESIDENTIAL ADDRESS. 

1899 and contributed five papers to the Proceedings. Though he resigned in 1933, he 
still maintained an interest in the Society and on numerous occasions rendered valued 
help by translating communications in foreign languages. 

Presidential Address. 

Pollen Development Patterns in the Epacridaceae. 

A Problem in Cytoplasm-Nucleus Interaction. 

Tetrad pollen is generally characteristic of the Order Ericales. In the tribe 
Styphelieae of the Epacridaceae this tetrad form has been altered by the regular failure 
of three pollen grains in each tetrad, to produce apparently single monad grains. In 
its most extreme form, monad development involves cytoplasmic polarity and nuclear 
migration in the mother cell following meiosis. This extreme monad form probably 
originated roughly contemporaneously with the origin of the tribe. In some genera 
and species it has suffered further modification, to give variable or segregating pollen 
tetrads and reversion to regular pollen tetrads. These patterns of pollen development 
are described and illustrated. An attempt is made to offer a general hypothesis to 
explain the mechanism of monad development and to follow up the evolutionary 
consequences of this hypothesis. 

(For full text see pp. 8-35.) 



The Honorary Treasurer, Dr. A. B. Walkom, presented the balance sheets for the 
year ended 28th February, 1959, duly signed by the Auditor, Mr. S. J. Rayment, 
F.C.A. (Aust.), and his motion that they be received and adopted was carried 
unanimously. 

No nominations of other candidates having been received, the Chairman declared 
the following elections for the ensuing year to be duly made: 

President: T. G. Vallance, B.Sc, Ph.D. 

Members of Coimcil: R. H. Anderson, B.Sc.Agr.; A. N. Colefax, B.Sc; J. W. 

Evans, M.A., D.Sc, Sc.D.; Dorothy A. Thorp, B.Sc; T. G. Vallance, B.Sc, 

Ph.D.; and J. M. Vincent, D.ScAgr., Dip.Bact. 
Auditor: S. J. Rayment, F.C.A. (Aust.). 

A cordial vote of thanks to the retiring President was carried by acclamation. 



BALANCE SHEETS. 
























en 


c^ 




13 










rH 












02 










00 






CO 

en 
00 


LO 




« 










CO 








C<1_ 

co" 

CO 
























«+< 




■d 


o 




CO ^ 









CJ2 






tc 


o 




t- ■* 









CO 








i-H 
























00 




Tf U5 









cj; 






trt 


o 
in 




4,84 
4,83 




1-1 




rH 








i-H 






^ 




















Ol -4-^ - 










. . 1 










he CO -^^ 








>. 




CQ 








^09 








a; 








CO 

o 
o 

4^ 




S " s 








CM 









as 




- ~ c^ "-^ 






















m -G 

















O 




Wat( 
ge B 
one-t 








c 








J 




CS ^ 








2 
3 






1 


,£3 




olita 
Drai 

ouse 




-2 


03 


ce 

pq 






0^ 


o 


CO 

0) 


g^^l 




CD 
CO 
CO 


K 







CO 


3 


"q; '^ "1^ 




<; 


c 


Sd 




< 


s 




g 


+- 
3 C^ 


3 


+; 




S2 




-« 


s 




.« 




s 


m 


c 






OJ 


o 


cp 









^ 


03 


[ 




X 


OQ 




VI 




^ 


00 


























fe 























-d 






^ 


Tt 








CO 


tr- 




M 






^ 










^ 


io 












T-\ 










rH 












t- 


t- 








02 


t- 










Ci) 


tc 








-*i 


r^ 




C4^ 






CO 


c 


^ 






00 


cq 


































-* 1 


-a 








rH 


CD 










C<l rH 1 


c 


> 








CO 












' 












tti 




-d 




o 











rH 


^ 



rH 






M 




o 
o 












CO 
00 


^ 

c 


CO 






trt 




o 
o 
^' 




CO 








CO 




LO 




CO 
























W 




s 




^ 
















1— ! 




.s 




'^ 
















)-H 






o 


CO 




















■^ 


s 




















QJ 


.0 




















s 

o 


a 




CD 

> 






C 
3 











m 


<M 


3 


CD 










^ 








-a 


■a 


■f^ 


a* 


CO 




CO 


<; 










3 


tp 


,— ; 


a; 


cp 




cu- 












> 


T3 


.^ 


tf 




*-»-^ 


M 


c 
c 








fe 


'a> 

o 


>» 


S 


CO 




15 


.3 


< 

a 






13 


OJ 


03 
OJ 


cc 


_CP 




03 


■^3 


(V 
m 




■g 


13 
03 




cu 




J 


.is; 


E 
c 


(P 




9 


3 


S 


ri^ 


M 




+-^ 





CO 




s 

3 


o 




-*— ' 


fl 




E^ 





^ 


^ 




S 




^ 






O) 


pq 




02 






<J 




fe 







3 










<1 













u 












1 



in 

0) 



s 



H 
2 

D 
O 

u 
u 

< 

% 

o 
o 
z 





-d 


00 








CTS rH 




00 rH 00 
ri 


CD CO 


to 


c^ 




cq c^3 o> cq 




ca 


'^ 05 00 ^ 




rH 




rH rH rH rH 




rH 


r-< 




^ 




cq 




oq cq 05 CO rH 00 




0> '^ rH 


Cq "# LO 1 Ir- 




10 




Cq rH Oq rH 




05 CO 




-f CO 1 00 


crt 


C-J 




"^l t- CO_ -* C-J 




•* 10 




rH 


io_ 




rH 




r-^ 
















10" 




























crt 


13 






0000 




















co" 






Cq CO -^ 
r^ rH 


























C^ t— ^ 






















crt 






cn cKi 

CO 












































■ • B 


CJ5 






































10 




































. w 


rH 




































. |3 












































































n 


>a 




































+J S-H 


^ 




































G 3 


2 




































03 ^ 






























































































LT 




















^« 


<p 
















I> 




















fc 
















LT 


















OJ 


















C 

a 


c/- 

c 
c 


8-59 . . . 
ears 

Advance . 
nciate 


CO 
CP 
CP 

0) 

-I- 

3 CO 


CD 
CO 



cp 




C 1^ 

CO 


00 

cq 

oi 
IP 


isferred) 
Expenses 
if Reprints 


ird Sales 
ng Grants 
y Sales . 




a: 


CA 


10 H I 

rH <; M < 


=* s p 


.u cr 


£ 




03 M 

^ fH CP 


stca 
inti 






"rc 






S rH C 


CP 'n 


. cp 






03 03 


%- 


■i^ 






CC 


a 




H wa 


Piv. 


z^ 






PP CO P^ pi^ CO 






> 


' 




- 




. 






- 




_ 








PC 


























^^ 






T-i 


CO 










1-H 


'^ 




CO 






rH 1-t 






















CO 






t~ C- 


rfl 
rH 










rH 


t= 


^ 






C\i 






7-i CO 


cq 










t- 


<= 





t~ 


-*l 






^ rH 













C<J 





cc 




00 


crt c~ 






00 


T-i 










tr- 


ic. 


_^CD 


LO_ 








rH" 
















' 




10 


■^ CO 





'J 


C^ oc 


CD OC 


CO c:: 


CTi 




C= 












CO 




CO 


t- CO 


CD C- t- 


cq r- 


00 «= 




c= 












rH 








T-i T-i T- 


r- 


















LO 


«: 


a- 


t- 10 


CO CT3 CD 


CO 


^ 




CD 












crt a> 


C3~ 


-* 


CD CO 


rH LO C= 


^ 


CD Oi 




l> 















-^ 


c^ 




rH r- 




rH rH 
















rH 


















ip 




































[3 




> 









































<P 

CO 








































Oi 




































CP 




p:; 




































»^ 








































l< 




CO 

<p 












W 






















bJO 


>•. 






























fi 









"£ 


CD 




tuO 






































c 













-S 


CO 












Vj 








CS 











CO 




















03 








P^ 


03 




Vj 












cp 


C 












-(-> 








»^ 




3 




03 








CJ 




S 


^ 










Id 




V. 






<=^ 









LO 
03 

rH 










o; c/ 




3 


p 


• 03 
4 CD 


a 


CL 


















3 
cr 


ranc 
age 
V C» 




M " 
s ^ 
Z! ^ 




t>. 3 

03 'S 


^ 


a 


<P 

CO 







CP 

a 






3 s 




c 
C 


Insu 
Post 
Pett 


'-3 S C 

3 E; X 


J 3 "^ 




_j-. 


1— 


p 


<; 








CC Ph Ph 


5 


<IJ^H 


opq 


3fe 


s 




H 




pq 






. 






. 


. 


_ 




., 




_ 










h 


























" 


' 




" 











pq 



M 



> cp -^ CO 





CM 



'3 ,a 
P5.S 


rH 


■a 
.-c 3 




>> 


0. ^ 


^ 


m 




.2 


> CP 


t>. 


3 




jr 


U 


<I3 




'0 


XJ OJ 


o3 









03 ^ 


3 .TH- 






CO 




^ ^ 


+^ 




s 


<P r 

rH (P 

' ' 


_D CP 


03 




oi 


15 iS 


+H 




<P 


0, 


3 




rH 


Xi CO 


3 




3 ^ s 


00 -rH 

cq 







CO 

Pi 


CP 


■^ ^ 


co-^ 


-3 

0) 


H 




cp .3 


^ 


J-. 


tw 





•rH <P 


(P 


pq 







Oj > 




§ 
H 


CO 




3 03 oj .d 

03 


3 


S 





O.-'g 


CO CO 


5 






> 


rH 


2 




Pi 

Ph 


03 




^1 


^H 

Pi 


Pi 




Cj 


CP S 
fc 


5 


C/2 


CO 



c^i 

£j_j 




cp 5^ 




tf 


S<^ 


s ° 






Fh 





^ 


to to 

CP 
cp +-> 




>— 1 

Q 


.a 




3 cS 




.^ 


x^ 


QJ |_| 


CD -*-^ 


, 


<i 


cp 


(P c 


rH !h 

5 <P 


o> 
Lc^ 




,r3 





o> 




^^ 


3 . 


rH 




TJ 


CP 3 


<P CO 


^ 




Oj 
13 


.c 2 


SO 


13 




S 
oJ 
X 

cp 


^ S 





03 

rH 




p> 


03 "O 


.^ 


rH 




S 


^g 


>^ 






'-', .^ 


>> fl 


CD 




HH 


*j (P 


S| 


C 






3 CP 


■3 






Xi 


•S^ 


>. 






CO CO 


CO 



BALANCE SHEETS. 



u 

< 

X 
H 
D 
O 
to 

z 

Lu 
O 

>■ 

H 
U 

u 
o 

z 
< 
u 
z 

z 



z 

O 

U d 
U in 



t/3 

a. 

CO 



O u, 

U, IN 



m 
<! 



J -, 



o 





o> 


o 


o 




-* 


lO 


o 




iH 


1—1 






to 


oa 


in 




la 


t- 


CO 




°\ 


iH 


o 












■to 


<m' 


to 




tH 






0) 


• 






bX) 








03 


+j 


+J 






CD 


M 




a> 


o 


O 




^ 


o 


O 




OJ 


,J_) 


) ^ 




02 


d 


03 




(1) 


03 


^' 






O 02 






W 


^ 


O) 


F^ 


OJ 




be 




M 


y-i 


03 




CS 


O 


bjo 



o 



fe 



— 03 .S .5 « 

OS .. t; "^ " £ 
^ 0- o Q m '^ 






i 




'O 


CS 





b 


OJ 


"£ 


OS 


^ 


^ 





ID 


fei 




rt 


0! 



OQ 








h-1 



o 



n( 


■« 


m 




Til 


>, 




0) 





C 


OJ 




a 


rr) 






C2 









u 





O CO 

o ^ 

O 00 



CO 
H 






03 ^ 

a; a 
;^ o 

C O 

^ l-l 



5 5 
o en 

S 3 



pq 



^ „. w 

03 00 O O iH 

tH r^ 

CO l^ O rH 

trt 00 ^ t~ 



tc 



p, 

5 to '-c +J 

s 1^ 3 a 

=t-l 1^ _S o 

CO o a; S ^ 

- S S S 13 

t. js a 3 -u 

i5 S iS O a 

oj 'D 03 OS 

M tf m o 

o - . 



<! 



S i^ rt '^ 



o3 



o> 



"S > £ b 

.rt O 5-1 h 

t:j ^ CD a 

° OS -^ => ^ 

- rH a> 4-> 

03 *^ Cl PH QJ 



;<1 



03 



_g^ OS 03 ^ 

-*J !h C3 Cfi r^ 

O) .,-1 !_ 

O -a OS > 

^H g o3 OS fl 



« <! 



H S 5 



o 05 sh oi a-" 
> '"' o S 



o '-^ 

^^ 

O " c 

S J3 ^ 

pi 
^ s ^ 






O 

1-3 



oj 






3 O) 

o xi 



ft >. 

O ,Q 

IS 

O 
fl Si 



U2 



BALANCE SHEETS. 



in 

u 

•J 
< 

X 
H 
D 
O 

u 
z 

u, 

o 

tH 

H 
u] 

U 
O 
en 

Z 

<: 

u 
z 
z 



in 

. 35 

O g 
< b 



o o 
o o 
oo cq 



02 



03 

C c 1 



-2 ts 



o 



m 






Ji 03 O 



a; o n 

S p -ci 

a; '^ 
Q 



03 
P< 

s 

O 

o 

M 

2 
S '^ 

03 -^ 



o 



<D 



9 C 

S 
o 
O 



o 


o 


o 


o 


o 


o 


o 
o 
o 


o 
I-l 

CO 

to 


o 



m 



>. 

.Q 



3 


equeath 
ay 

ted Inco 
Fund . 






nt b 
acle 
lula 
rch 


P 


P g S 03 




o P a) 






3 




<: <i1« 



03 





•6 


o 


o 






o 




lA 


CO o 
t- o 

(M to 






CO 

c- 

00 




crt 


iH CO 






c~ 














crt 


0) 














05 




00 ■ 
U5 








1-H 
















05 








>^ 




tH 








fa 












l« 




^ 








s 




c ■ 








fa 




p 








^ 












« 




d 








u. 




Oj 1-i 








J3 




o3 d 








00 










N 












TS 










U 












no 




>. 








e 




CQ 








u 












fa 












It 

4) 


-d 


■^ 


lis 


t- 


o 


tc 


c 


CO 


02 


CO 








iH 


iH 


r-l 


H 




c 




to 


C~ 


Z 




<z 




00 


00 


« 


c— 




^ 


O 












U 












o 




i 








<: 





J 








u 












§ 




CM 

o 








o 




^ 








o 




;h 








z 




,2 












02 

■a 

O 

>. 
CD 

d 
o 




LT 


) 
i 

H 






>! 


a 
^ 1 


C 


) 






'm t 


H C 


a 


} 






t^ a 


IJ o 


1 


1 






<D J 




e 








> : 


3 - 


) 




















C 





i 






P 


JP 


5 






o 




. 


^ 






H 




" 


*" 



>d H 



m 



W 





fe 0) -^ CO 

?> "^ ^ c3 




U 




>5f C 3 

^1 cr; 




o 






A 


■3 
d 




>. CD ?: - 


4.J 
CO 




ociet 
abov 
here 
uary 


p 






< 






m -^ !- -a 


■4-) 






ean 
t the 
ance 
Feb 
pecte 


d 






n o3 -rt .o M 


d 






CJ 43 si. -^ Cl 


o 






•S +j o 2S •'^ 


o 






J ^ o"^ 


o 




!/2 


£? " -M « 
0) .-a 0! 15 0) 


<5 




tf 


















en's rt W 03 
2 g tS rt ^ 

(11 TO 


o3 
.d 




o 


vouchi 
1959, 
;orrect 
ciety's 
ments 


a 




Pi 
o 


and 

uary, 
are c 
le SO' 
nvest 


< 

1-5 




p4 


of account 
28th Febr 

e Account 
state of tl 

!S of the i 




Pi 
m 

o 


m 




H 


Dks 

ded 

comi 

rue 

cate 






s 






u 


5S£:§ 






< 










cri 




<i> ^ M :3 Oj 




in 




^ 0) fi O 




o> 








r^ 




■^ CD d 0) m 




d 

p 

S 
^ 




examine 
es for th' 

accompa 
lion pres 
the booki 








4-> 




lave 
Wal 
and 
opir 

by 




1— 1 








>^ 




. . ^ ■« >> d 




0) 




1 
Sout 
Shee 
in m 
show 




d 
■a 

CO 



PRESIDENTIAL ADDRESS. 



POLLEN DEVELOPMENT PATTERNS IN THE EPACRIDACEAE. 

A Problem in Cytoplasm-Nucleus Interaction. 

By S. Smith-White. 

(Plates i-ii; twelve text-figures.) 

[Delivered 25th March, 1959.] 

Synopsis. 
Tetrad pollen is generally characteristic of the Order Ericales. In the tribe Styphelleae 
of the Epacridaceae this tetrad form has been altered by the regular failure of three pollen 
grains in each tetrad to produce apparently single monad grains. In Its most extreme form, 
monad development involves cytoplasmic polarity and nuclear migration in the mother cell 
following meiosis. 

This extreme monad form probably originated roughly contemporaneously with the origin 
of the tribe. In some genera and species it has suffered further modification, to give variable 
or segregating pollen tetrads, and reversion to regular pollen tetrads. 

These patterns of pollen development are described and illustrated. An attempt is ipade 
to offer a general hypothesis to explain the mechanism of monad development and to follow 
up the evolutionary consequences of this hypothesis. 



The interests of this Society are very diverse, and over the years presidential 
addresses have covered an equally diverse field. Many have presented reviews of the 
fields of knowledge and research in which the president of the year has had standing. 
It is not my intention to follow this custom. Instead, I propose to present and discuss 
a problem peculiar to a group of AustraUan plants which has interested me for many 
years but on which there is practically no published information. 

My story concerns the characteristics of mature pollen and the patterns of pollen 
development found in the tribe Styphelieae of the Epacridaceae, and it involves 
differentiation at the intracellular level. Problems of differentiation provide one of the 
central themes of present-day biology, and I believe you will find my story unusual 
and interesting. I cannot provide the answers, but I am prepared to do some guessing. 
It has been said that the tortoise can make no progress until he sticks out his neck! 
I will become involved not only in cell polarity and cytoplasmic gradients, and the 
immediate reactions of nuclei to these cytoplasmic conditions, but also, in the long 
term, in the consequences of cytoplasmic polarity on nuclear genotype. 

The Por.iEN or the Styphelieae. 
In the Order Ericales, and particularly in the major family of the Order, the 
Ericaceae, the pollen is characteristically matured in tetrads. Only in the monogeneric 
family Clethraceae do the individual pollen grains separate during maturation. In 
the Epacridaceae tetrad pollen is also produced, and this form is, to my knowledge, 
uniformly characteristic of the tribe Epacrideae, In the Styphelieae, however, there 
is an unusual diversity of pollen types. 

The Monad Pollen Type. 
As early as 1810 Robert Brown observed that in this tribe the pollen may be 
either tetrad or single. Brough (1924) described the pollen of StypJielia longifolia 
as being single. In my experience, truly single pollen grains do not occur within the 

Proceedings of the Linnean Society* of New South Wales. 1959, Vol. Ixxxiv, Part 1. 



BT S. SMITH-WHITE. 9 

tribe. Each apparently single pollen grain is a tetrad containing one good pollen 
grain and three aborted and empty cells, all enclosed in a common thick exine. The 
terms "pseudomonad" and "monodynamosporous tetrad" have been used for the 
somewhat similar pollen type found in the Cyperaceae (Selling, 1947; Erdtman, 1952), 
but they seem unwieldy and unnecessary. The term monad is sufficient to distinguish 
this type from single pollen grains and will be used in this address. 

The main features of the development of monad pollen in Styplielia and in 
Leucopogon have been described in the Proceedings of this Society and elsewhere 
(Smith-White, 1948a, 1955a, 1955b). In the pollen mother cell (PMC) immediately 
following meiosis, the four microspore'^ nuclei are uniformly spaced, either in tetra- 
hedral or quadrant arrangement (Text-fig. 1, a). Before partition of the mother cell 
cytoplasm occurs, migration of the nuclei takes place; three cluster at one end of the 
cell and the fourth moves to, or remains in, an isolated position at the opposite end 
(Text-fig. 1, h). Cell walls are then laid down and the microspores are constituted. 
These microspores are initially very unequal in size (Text-fig. 1, c) and unlike in 
potential. The differentiation of the one functional and the three non-functional micro- 
spores occurs at a very early stage, either immediately following or perhaps during 
meiosis. All four microspores undergo some development, and there is some thickening 
of the cell walls surrounding the small ones. Very soon, however, the three minor 
microspores die (Text-fig. 1, cl, e). As Barber has shown in Uvularia (1941) the 
opportunity for intercellular compensation is then lost and the small microspores can 
derive no advantage from their association with the large one to compensate them for 
their own deficiencies. 

Subsequent growth of the tetrad and its increase in size with maturation is entirely 
due to the large microspore. The aborted cells become compressed and obscured 
(Text-fig. 1, g, h). In the functional microspore the pollen grain mitosis occurs at a 
rather late stage of development, and after very considerable increase in size. This 
PG mitosis shows an inversion of the usual polarity characteristic of the PG mitosis 
in many fiowering plants (Sax and Husted, 1936; Brumfield, 1941; La Cour, 1949; 
de Almeida and Sampaya, 1950), in that the short pole of the mitotic spindle is 
directed radially outwards from the centre of the tetrad, and the generative nucleus, 
and later the generative cell, is derived from this outer pole (Text-fig. 1, /, g). Finally, 
the generative cell migrates into the centre of the pollen grain and its boundary 
becomes indistinct (Text-fig. 1, h). 

This pattern of pollen development is the most frequent one in the tribe. It has 
been termed the Styphelia type, or S-type, since it was first found and is most fully 
developed in that genus. It is not usually associated with visible meiotic abnormalities. 

Normal tetrad pollen (T-type), similar to that of the Ericaceae and the Epacrideae 
is also found in the Styphelieae, but is much less frequent. It is known to me in 
11 species in the genera Acrotriche, Cyathodes, Pentachondra. Trochocarpa, Leucopogon, 
Brachyloma and Lissanthe (Table 8). Three other pollen types are also found in the 
tribe and involve patterns of development intermediate between the S-type and the 
T-type, in a sense linking them together. These will be described and considered after 
the implications of the S-type monad pattern of development have been discussed. 

The Implications of the Styphelia Pattern of Development. 

The unlike fates of the microspore nuclei in the Styphelia mother cell demonstrate 
the existence of differentiation within the cell. This differentiation may be in the 
cytoplasm or in the nuclei, or more probably in both. 

Intracellular polarity, i.e., the development of substantive gradients between 
different parts of the cell cytoplasm, is not infrequent and is perhaps a necessary 
concomitant of all primary cellular differentiation (Mather, 1948). Such gradients 

1 1 propose to use the term microspore for the four daughter cells of a PMC until they 
reach the stage of the first gametophytic mitosis, the pollen grain mitosis. After this mitosis 
they become binucleate pollen grains. 



10 



POLLEN DEVELOPMENT PATTERNS IN THE EPACRIDACEAE, 



are evident in animal eggs, in blood precursor cells (La Cour, 1944), and in pollen 
grains (Geitler, 1935; Ostergren, 1947; La Cour, 1949). Similar gradients occur in 
embryo sacs (Darlington and La Cour, 1941; Darlington and Mather, 1949; Brock, 
1954) and probably control the various patterns of embryo sac development which 
have been described and illustrated by Maheshwari (1950). 




Text-fig. 1. — Pollen development in Styphelia longifolia. x ca. 1150. 



Very often intracellular differentiation can be related to the existence of physio- 
logical or tissue gradients in the surrounding tissues or, more generally, to a polarity 
of the surrounding environment: intracellular polarity is imposed from outside. Such 
is not always the case. La Cour (1949) and de Almeida and Sampaya (1950) have 
shown that the orientation of the pollen grain mitosis and the differentiation of the 
generative and vegetative nuclei are determined in relation to the centre of the 
original pollen mother cell. Each mother cell must be a centre of development of 
polarity. In the monad pattern of development each mother cell also seems to be a 
centre of origin of differentiation. There is a random orientation of the monads to 
the anther axis, and no general tissue gradient can be inferred. The monads are not 



BY S. SMITH-WHITE. 



11 



arranged in opposite pairs, as would happen if differentiation originated at the 
premeiotic mitoses. The control of monad development must be intracellular and 
local; in fact the usual and normal pollen grain polarity in many angiosperms and 
the exceptional polarity in the pollen mother cells of the Styphelieae responsible for 
monad development, are so similar in action that the former might well be the source 
of the latter. A slight precocity in the development of the PG polarity, bringing it 
forward only one cell generation, may be the fundamental change involved in the 
origin of Styphelia type pollen. Two examples of a greater precocity, causing the 
Intrusion of polarity into the premeiotic mitosis, have been reported by La Cour (1949) 
and by Holden and Mota (1956). 




Text-fig. 2. — Two hypotheses of monad development. 
^STuclear (chromiosomal) determination. 



A. Cytoplasmic determination. E. 



There seems no alternative to the view that cytoplasmic polarity is involved in 
the control of the monad pattern of pollen development, but it may not be sufficient. 
In dealing with a possible case of polarized segregation in Bomharclia, Catcheside 
(1944) has reasoned that it must involve both (i) a polarity in the cytoplasm or in the 
spindle, and (ii) some difference between the genes, which show a differential 
response to this polarity. His reasoning is equally valid where the units are whole 
chromosomes or whole nuclei. In monad development this nuclear differential could 
consist merely in the spatial arrangement of the four microspore nuclei or of the 
meiotic spindles in relation to the cytoplasmic axis. With the usual tetrahedral or 
quadrant arrangements one nucleus would often lie closer to the "effective" cytoplasmic 
pole than the other three, so that it might outgrow and suppress its sisters. This 
liypothesis is illustrated in Text-figure 2-\. 

The hypothesis needs a further brief consideration. Two features suggest that it 
is incomplete. It might be expected that occasionally two of the four nuclei would 
1)6 approximately equally favoured in position so that dyads would be produced. 



12, . POLLEN DEVELOPMENT PATTERNS IN THE EPACRIDACEAE, 

However, dyads have not been seen in any species of Styphelia, and in Leucopogon 
juniperinus only one dyad has been seen among over 300,000 monads. The system of 
control is extraordinarily efficient. La Cour (1949) has shown that in Scilla mis- 
orientation of the pollen grain mitotic spindle in relation to the cytoplasmic polarity 
may be brought about by experimental shock treatment, and that this misorientation 
leads to a failure of nuclear differentiation. The second feature which calls for an 
addition to the hypothesis is the occurrence of nuclear migration. Can a qualitative 
difference in nuclear response arise as a result of a quantitative difference in the 
concentration of some cytoplasmic activator? This problem also exists in the normal 
pollen grain differentiation and it emphasizes the similarity of the two systems. 

Both of these features can be satisfied by a simple addition to the hypothesis. A 
spatial orientation of the division spindles in the cytoplasmic gradient could lead to 
an unequal distribution of non-chromosomal or non-genic nuclear materials and thus 
to a constitutional difference between the four nuclei. This constitutional difference 
could then condition the qualitative difference in response and the system would 
become self-reinforcing and reliable. 

A second hypothesis is presented in Text-flgure 2n merely to be eliminated. It 
involves chromosomal segregation and requires very precise and unlikely conditions. 
It requires the segregation of two loci of lethals A/a and B/b, which must be placed 
on the same chromosome pair at such a distance that one chiasma, and one only, is 
always formed between them. It thus requires chiasma localization. It also requires 
the maintenance of permanent hybridity by a system of complementary survival of 
AB pollen grains and ab embryo sacs. These conditions are unlikely. Again, it would 
be expected that such a system would break down in polyploids, but monad pollen 
is characteristic of both diploid and polyploid species, and even of the permanent 
triploid Leucopogon. The monad system breaks down only in secondary polyploids. 
The most serious objection to this second, segregational hypothesis, however, is that 
it is independent of cytoplasmic conditions and renders the assumption of cytoplasmic 
polarity superfluous. Thus it cannot explain nuclear migration and it isolates the 
pattern of monad development from the very similar patterns of differentiation in 
pollen grains and embryo sacs, where segregation is not possible. 

At the present stage the first hypothesis must be favoured. Monad development 
is related to pollen grain differentiation and is determined and controlled by a 
cytoplasmic polarity and a consequent non-chromosomal nuclear differential, dependent 
upon the spatial relationship of the four microspore nuclei to the cytoplasmic polarity. 
However, with the intrusion of cytoplasmic polarity into the mother cell during meiosis, 
cytoplasmic control of segregation becomes a logical possibility, provided heterozygosity 
is present and provided that permanent heterozygosity can be maintained. In this 
connection the case of Leucopogon juniperinus is particularly instructive. 

The Special Case of Leucopogon junijje^'inus. 

This species is a permanent triploid numerical hybrid belonging to a section of 
the genus in which the haploid number is four. The pollen is similar to that of 
Styphelia and is of the extreme S-type, with marked nuclear migration occurring in 
the pollen mother cells. The somatic chromosome number in the species is 12, and 
at meiosis four bivalents are regularly formed, leaving four chromosomes as univalents. 
In the pollen mother cells the univalents show polarized segregation, all passing to 
the same pole with a frequency greatly in excess of binomial expectation (Smith-White, 
1948). This polarized univalent behaviour is apparently present also in the embryo 
sac mother cell meiosis (Smith-White, 19556). Permanent hybridity is maintained by 
complementary gametic survival, whereby the univalents are regularly excluded from 
the functional pollen and equally regularly are included in the functional embryo 
sacs. This system is illustrated diagrammatically in Text-flgure 3. 

A similar controlled and polarized segregation of univalents occurs in the flrst 
meiotic divisions in embryo sac mother cells in Rosa caniyia (Tackholm, 1922; Hurst, 
1931; Gustafson, 1944). In this species complex there are seven bivalents which show 



BY S. SMITH-WHITE. 



13 



normal segregation, and 21 or 28 univalents all of which pass to the micropylar pole 
of the embryo sac mother cell at Anaphase 1 and are thus included in the embryo 
sacs. The system of Rosa canina, which has been called semi-apomictic by Darlington 
(1939), differs from that of Leucopogon juniperiniis in that in Rosa canina there is no 
control of univalent segregation in the pollen mother cells. 

There can be no question or doubt that the S-type of monad pollen is antecedent 
to the condition of triploidy in Leucopogon junipei-inus . The species yields evidence 
of fundamental importance for the analysis of the problem of monad development. 

(1) It shows that the monad pattern need not be upset by triploidy. Therefore, 
chromosomal segregation is unlikely to be a primary cause of the differentiation. 

(2) It emphasizes the presence in the pollen mother cell of a cytoplasmic gradient 
and it defines the time of action of this gradient, which must be developed during 
or before the metaphase of the first meiotic division. Polarity operates at a time 
when it could influence the segregation of unbalanced chromosomes. 




P.M.C. 



and 



POLLEN GRAIN 



E.SM.C. and EMBRYO SAC 




Text-fig'. 3. — The complementary gametic system in L-eucopogooi juniperioius. The four 
univalents are regularly included in the embryo sacs, and excluded from the pollen. 



(3) It demonstrates that unbalanced chromosome univalents can and do react to 
the cell polarity, whereas paired chromosome bivalents, which are subject to the much 
stronger forces of centromere repulsion, are not free to do so. 

Swanson (1958) has discussed the literature dealing with preferential chromosome 
segregation and a repetition is not necessary here. A short account of cases which 
may be relevant to our problem will suffice. 

There is some evidence that preferential segregation is often a function of the 
centromere or of heterochromatin in the centromere region. This is true of trisomic 
segregation in triplo-IV D?'osophila melanogaster (Sturtevant, 1936) and of X chromo- 
somes in the same species (Novitski, 1951, 1954; Sandler and Braver, 1954) in Sciara 
(Metz, 1938; Grouse, 1943) and in B-chromosomes in Maize (Roman, 1947a, 1947&). 
In Maize heterozygous for a terminal knob on chromosome 10 (Rhoades, 1942, 1952) 
preferential segregation, leading to the inclusion of the knobbed chromosome 10 in 
the functional (chalazal) megaspore, is associated with neocentric activity of the 
abnormal heterochromatic knob. It is therefore significant that clonally-inherited 
chromosomes or chromosome segments, such as the univalents of Rosa canina and 
Leucopogon juniperinus, B-chromosomes, and the differential segments of sex chromo- 
somes, are very often heterochromatic. 



14 POLLEN DEVELOPMENT PATTERNS IN THE EPACEIDACEAE, 

The controlled segregation of XYiY, or of X^XoY sex chromosomes, which is 
known in Humulus japonicus (Kihara and Hirayoshi, 1932), in Rumex acetosa (Kihara. 
and Ono, 1925, 1928) and in several groups of insects, including some Orthoptera^ 
Dermaptera and Heteroptera (lists and references given by Darlington, 1937, and 
White, 1954), is determined by the properties of associated centromeres (Darlington, 
1937, p. 362) and is not relevant to our problem. The segregation of sex chromosomes 
in sex mechanisms of the XiXoO type found in spiders (White, 1954) and the 
X1X2X3X4O type in Euceraphis, where all the X chromosomes pass to the same pole, 
and the co-orientation of X chromosomes at M2 in Cimex (Darlington, 1940) cannot be 
explained in the same way.* Swanson (1958) does not suggest an explanation for 
such behaviour, but a mechanism described by Catcheside (1950) in Parthenium 
argetatuni could be operative. In this species, heterochromatic B-chromosomes show 
secondary association during meiotic prophase. Thus they tend to approach the 
metaphase I plate from the same direction, lie all on the same side of the 
A-chromosome bivalents, and pass together to the same anaphase pole. 

Catcheside's mechanism implies preferential but not polarized segregation. It is 
not sufficient to explain univalent behaviour in Rosa canina embryo sac mother cells 
or in Leucopogon juniperinus pollen mother cells. It could, however, suffice to explain 
univalent segregation in embryo sac mother cells of the latter species, since the four 
univalents move together to either the micropylar or chalazal poles, apparently with 
equal frequency. In the embryo sac mother cells of Rosa canina and the pollen mother 
cells of Leucopogon juniperinus an additional mechanism is necessary to direct the 
univalents to a particular pole. Since in both there is other evidence of an 
intracellular cytoplasmic gradient, it is not unreasonable to suggest that this cytoplasmic 
gradient is the determining cause. The Parthenium. behaviour then becomes superfluous, 
and in fact secondary association of the univalents of Leucopogon juniperinus during 
prophase has not been observed. 

(4) Finally, Leucopogon juniperinus demonstrates that the same system of intra- 
cellular differentiation operates in both the pollen mother cells and the embryo sac 
mother cells in respect to univalent segregation, but that the consequences, the 
elimination or retention of the univalents in the pollen grains and in the embryo sacs, 
are different (Smith-White, 1955b). This difference must be dependent upon the 
different tissue environments of the anther and ovule. A surrounding tissue polarity 
is present in the ovule and is evidenced by the development of the micropylar 
megaspore rather than the chalazal one. A tissue gradient is not present in the anther. 

The morphological uniformity and widespread but discontinuous distribution of 
this triploid species suggest that it is of very considerable age. In the course of 
time the univalents, subject to purely maternal inheritance, may have become vital in 
embryo sacs and lethal in pollen, but they cannot provide the primary cause of the 
monad pattern of development. The opposite contention, that the system which 
determines the S-pattern also influences univalent segregation, would, however, seem 
to be valid. The conditions which determine complementary gametic survival and 
which stabilize hybridity in the triploid are also present in diploid species; there is in 
fact pre-adaptation to permanent hybridity. Conditions which provide an opportunity 
for the reinforcement of a cytoplasmic control of the monad pattern of development 
by a chromosomal or segregational mechanism are demonstrable in the triploid and 
can be inferred in the diploids. 

Modification of the Monad Pattern of Development. 
Some species of Leucopogon (e.g. L. revolutus, Smith- White, 1955o) produce monad 
pollen which, when mature, is scarcely distinguishable from the S-type already 
described. In development, however, there is an absence of nuclear migration following 
meiosis, so that the four microspores are initially of equal size. Despite this apparent 
equality, three microspores in each tetrad regularly die at a very early stage of 
development. The term modified monad and the abbreviation S'-type will be used to 
designate this modification and the abbreviation M-type will be used to include both 
S-type and S'-type. 



BY S. SMITH-WHITE. 



15 



The S'-type of monad development is associated, so far as is known, with regular 
meiosis. It is still necessary to infer the existence of an intracellular cytoplasmic 
differentiation to account for the different fates of the four microspores and for the 
stability of the system, even though there is no direct evidence of cell polarity affecting 
nuclear movement. The distinction between the S- and S'-types is in fact quantitative 
rather than qualitative. In L. ericoides there is evidence of nuclear migration, but it 
is of lesser degree than that usual in Styphelia, and the initial size-inequality of the 
four microspores is also of lesser degree, although still quite distinct. In Melichrus 
urceolatus there is variation within anthers, some mother cells showing marked nuclear 
migration and others showing little or none. 

This quantitative intergradation between the S-type and S'-type of development 
suggests that the control mechanism involved in monad development is itself quanti- 
tative, and it makes any classification of species into two classes, with S-type pollen 
and S'-type pollen respectively, such as was attempted earlier (Smith-White, 1955a), 
difficult and arbitrary. 

Breakdown of the Monad Pattern. 
It might be anticipated that the S'-type of pollen development would involve a 
lesser degree of cytoplasmic differentiation than the extreme S-type and that it might 
therefore be more easily subject to breakdown. This is the case. Comment has 
already been made upon the remarkable stability of the S-type system in Styphelia 
and in Leucopogon juniperinus. By comparison, Leucopogon virgatus (n = 10), which 
has S'-type pollen, shows a relative instability, with occasional or even frequent 
breakdown of the monad system leading to the formation of dyads and very 
exceptionally of triads. In a population of this species at Rylstone, N.S.W., 50 plants 
have been examined. Of these, many show no breakdown or a negligible degree of 
breakdown with less than 0-1% of dyads. Others produced dyads with a frequency of 
1-2% in every anther and two have a breakdown frequency exceeding 10%. Data 
from eleven selected plants of this population, presented in Table 1, illustrate the 
range of behaviours. 

Table 1. 

Breakdown of Monad Type Pollen in L. virgatus R.Br, at Rylstone, N.S.W. 

Tetrad-type Proportions. 







Percentage Frequency of Tetrad Types. 






Ratio 


Plant. 








N. 


Dyads/ 














Os. 


Is. 


2s. 


3s. 


4s. 




Monads. 


57/4 


37-85 


62-15 


0-00 


0-00 


0-00 


1638 


, 


57/11 


0-75 


99-25 


0-00 


0-00 


0-00 


1196 


— 


57/7 


2-55 


97-41 


0-04 


0-00 


0-00 


2240 


0-00045 


57/3 


3-08 


96-73 


0-19 


0-00 


0-00 


3244 


0-0019 


58/56 


2-66 


97-05 


0-29 


0-00 


0-00 


13634 


0-0022 


57/2 


6-24 


93 - 34 


0-42 


00 


0-00 


2371 


0-0045 


57/1 


4-68 


94-34 


0-98 


0-00 


0-00 


4810 


0-0103 


58/94 


0-61 


96-04 


3-35 


0-00 


0-00 


9586 


0-0349 


58/73 


39-17 


57-53 


3-30 


0-00 


0-00 


3360 


0-058 


57/10 


3-01 


87-28 


9-67 


0-04 


0-00 


2625 


0-111 


58/8 


1-37 


88-46 


10-06 


0-11 


0-00 


4740 


0-114 



Several features in this breakdown of the monad type need to be emphasized. In 
some plants there is produced an appreciable frequency of nullads (i.e. completely 
aborted tetrads, in which the usual single pollen grain has failed to develop), but 
high nullad frequency does not appear to be associated with either high or low dyad 
frequency. Secondly, breakdown leading to dyad formation is not an all-or-nothing 
process. When a dyad is formed instead of the more usual monad, the two developed 
grains in the dyad may be equal in size and both either apparently viable or partially 
degenerate, or they may be slightly or grossly unequal (Text-fig. 4, a-j; Plate i, 
figs. 1-12). All degrees of breakdown, from the perfect monad to the full dyad, occur 



16 



POLLEN DEVELOPMENT PATTERNS IN THE EPACRIDACEAE, 



in the same anther. The classification into monads and dyads given in Tables 1 and 2 
thus involves an arbitrary decision of the degree of development which should be 
scored. In the data, grains showing development of a second cell equal to or greater 
than that illustrated in Text-figure 4, f, have been scored as dyads. 

It will be observed that there are two different quantitative aspects of breakdown. 
These are (1) variation in frequency and (2) variation in degree. Breakdown does 
not involve a switch from one developmental pathway to another, and the control 
system of monad development must be itself quantitative. The behaviour supports 
the inference which was made earlier, i.e. the hypothesis that monad development is 
dependent upon intracellular cytoplasmic differentiation and polarity. 




Text-fig-. 4. — Monad breakdown in Leucopogon virgatus. x ca-. 1150. 
monad, c, d. Degenerate dyads, e-h. Degrees of monad "breakdown". 



o. Nullad. 
i-j. Dyads. 



b. Normal 



In Tables 2a and 2b data are given for individual flowers and anthers, to show the 
extent of variation in the frequency of breakdown within plants. It is clear that 
significant differences occur, between flowers and even between anthers of the same 
flower, and such differences must be environmentally determined. A comparison of 
Tables 1 and 2 shows that intraplant differeiaces are of a substantially lower order 
than are interplant differences. It may be inferred that there are genotypic differences 
between plants in the population affecting the stability of the monad developmental 
system and that where the system is weak, buffering against environmental breakdown 
is marginal. 

The precise nature and mode of action of environmental factors on monad 
breakdown remain to be determined, but they will probably be susceptible to 
experimental study. It is probably significant that there can be a substantial lapse 
of time, perhaps as much as a week, between the occurrences of meiosis in the five 
anthers of one flower, and that meiosis may commence at almost any time of the day 
or night. Thus it is possible that temperature at the time of meiosis or at some 
other critical stage of development could be an important factor in breakdown. 



BY S. SMITH-WHITE. 



17 



Table 2a. 
Variation in Monad Breakdown within Plants. L. virgatus R.Br., RyUtone, N.S.W. 
Variations between flowers within plants. 58/56, 58/73, 58/94 and 58/8 (triads omitted). 



Flower. 


Plant 58/56. 

Percentage Frequency of 

Tetrad Types. 


N. 


Flower. 


Plant 58/94. 

Percentage Frequency of 

Tetrad Types. 


N. 




Os. 


Is. 


2s. 


Os. 


Is. 


2s. 




1 
2 
3 


2-57 
2-43 
3-02 


97-32 
97-05 
96-78 


0-11 
0-52 
0-21 


4508 
4783 
4343 


1 
2 
3 


0-58 
0-56 
0-70 


96-75 
95-87 
95-58 


2-67 
3-57 
3-73 


2955 
.3195 
3436 


Total .. 


2-66 


97-05 


0-29 


13634 


Total . . 


0-61 


96-04 


3-35 


9586 


Contingency x' = 18-40. P<0-001. 


Contingency x'' = 6-78. P=015. 


Flower. 


Plant 58/73. 

Percentage Frequency of 

Tetrad Types. 


N". 


Flower. 


Plant 58/8. 

Percentage Frequency of 

Tetrad Types. 


N. 




Os. 


Is. 


2s. 


Os. 


is. 


2s. 




1 
2 
3 


42-20 
42-66 
32-25 


54-49 
54-27 
64-22 


3-31 

- 3-08 

3-53 


1147 
1137 
1076 


1 
2 
3 


0-85 
1-82 
1-52 


86-84 
93-16 
86-50 


12-31 

5-02 

11-97 


1649 
1374 
1712 


Total .. 


39-17 


57-53 


3-80 


3360 


Total . . 


1-37 


88-55 


10-07 


4735 


Contingency x' = 32-03. P<0-001. 


Contingency x' = 59-37. P<0-001. 



Table 2b. 

Variation in Monad Breakdown within Plants. L. virgatus R.Br., Ryhtone, N.S.W. 
A^ariations between anthers and flowers in 58/8. 



Flower. 


Anther. 


Percentage Frequencies of Tetrad Types. 


Total. 
















Os. 


Is. 


2s. 


3s. 


4s. 






1 





93-30 


6-42 


0-28 


. — 


358 




2 


1-90 


88-08 


9-76 


0-27 


— 


369 


1 


3 


0-77 


82-26 


16-97 





— 


389 




4 


0-62 


90-0 


9-38 





— 


160 




5 


0-80 


81-96 


16-71 


0-53 


— 


377 




Totals . . 


0-85 


86-63 


12-28 


0-24 


— 


1653 




1 


2-83 


88-68 


8-49 





. — 


318 




2 


2-01 


94-63 


3-02 


0-34 


— 


298 


2 


3 


0-60 


94-89 


4-50 





— 


333 




4 


0-65 


96-13 


3-23 





— 


155 




5 


2-58 


92-62 


4-80 





— 


271 




Totals .. 


1-82 


93 09 


5-02 


0-07 


— 


1375 




1 


1-28 


83-89 


14-83 





— 


391 




2 


0-50 


87-69 


11-80 





— 


398 


3 


3 


0-81 


84-84 


14-34 





— 


244 




4 


2-24 


85-62 


12-14 





— 


313 




5 


2-73 


89-89 


7-38 





— 


366 




Totals . . 


1-51 


86-50 


11-97 





— 


1712 


Grand totals 




1-37 


88-46 


10-06 


0-11 


— 


4740 



Neglecting triads, contingency, total Xas =137-6. P<0-001. 

between flower totals x. = 59-37. P<0001. 



is 



POLLEN DEVELOPMENT PATTERNS IN THE EPACRIDACEAE, 



Variable Pollen Tetrads. 

Breakdown of the monad system connects the S- and S'-types with still another, 
that of segregating or variable pollen tetrads. This type was first reported in 
Astroloma ptnifoliuni (Smith-White, 1948a) and has been designated as the Astroloma 
type or A-type. Similar segregating tetrads were later reported in Astroloma 
conostephioides, Brachyloma py-eissii. and Acrotriclie fascicuUflora (Smith-White, 1955a). 
Since then the type has been found in eight additional species in Acrotriche, 
Brachyloma, Leucopogon and Lissanthe (cf. Table 3, in which information is also 
given on chromosome numbers and on the pollen types of other species of the same 
genera). 

In A-type pollen development meiosis is followed by the partition of the pollen 
mother cell into four apparently equal microspores. All, any or none of these may 

Table 3. 

Species of the Siyphelieae showing Tetrad Segregation. 

A-type Pollen. 













Characteristics of 


Genera and 
Species. 


Number 

of 

Plants 

Examined. 


Number 

of 
Plants 
with 


Chromosome 

Number 

n. 


Associated 

Meiotic 

Irregularities. 


Congeners 




n. 


Number 
of 


Pollen 






A-type. 








Spp. 


Type. 


Astroloma 










4 


14 


S 


pinifolium Benth. 
conostepldoides F. Muell. . . 

Acrotriche 


>100 
>100 


All 
All 


7 


Fragmentation 
Inversion 
hybrid! ty 


9 


3 


T 


cordata E,.Br. 


3 


3 


Not known 










fascicuUflora Benth. 


20 


All 


9 


Hetero- 
chromatin 








divaricatu R.Br. . . 


5 


1 


9 










depressa R.Br. 
Brachyloma 


2 


2 


Not known 




9 


1 


S,T 


ericoides Sond. 


3 


3 


Not known 










preissii Sond 


6 


6 


" 


Possible 
fragmentation 








scortechinii F. Muell. 


1 


1 


Not known 










Lissanthe 










7 


1 


T 


montana R.Br. 


10 


2 


14 


Gynodioecism 








strigosa R.Br 

Leucopogon 


50 


1 






4, 6 
7. 10 




S, S' 
T 


melaleucoides A. Cunn. 


15 


15 


Not known 


Gynodioecism 









then abort, yielding five categories of tetrad in the mature anther — nullads, monads, 
dyads, triads and full tetrads (Text-flg. 5; Plate ii, figs. 1-6). 

Within genera, A-type pollen is associated either with the S- and S'-types or with 
the full tetrad T-type. Unfortunately, chromosome numbers are not known for all 
species, but so far as is known the A-type is not associated with the haploid numbers 
of 4 or 6. In four cases it is associated with abnormal meiotic conditions. 

The details of pollen development subsequent to meiosis in the twelve species listed 
differ only in minor respects. The following description, however, applies particularly 
to Astroloma pinifolium. 

Following meiosis the microspore nuclei assume a tetrahedral or a quadrant 
arrangement within the mother cell. There is no nuclear migration and the four 
microspores are initially equal in size. Inequality becomes apparent only after the cell 
walls separating the microspores have become appreciably thickened, but before there 
is any considerable increase in size, and is evident in the degeneration of the cytoplasm 
in some cells (Text-flg. 6, a-h) . As in the development of monad pollen, aborting 



BY S. SMITH-WHITE. 



n 



microspores fail after the opportunity for intercellular compensation has been lost. It 
is not clear whether the deficiencies which lead to death are nuclear and chromosomal, 
as in Uvularia (Barber, I.e.), or whether they are cytoplasmic and related to the 
gradients involved in S-type monad development. 

The pollen grain mitosis occurs in surviving microspores after these have under- 
gone considerable growth and is similar to the pollen grain mitosis in Styphelia in 
respect to the reversal of generative-vegetative pole polarity as compared with Scilla 
and Luzula (Text-fig. 6, g, h). 

In mature pollen, aborted cells are very small and completely devoid of cytoplasm. 
In some plants, however, a second kind of degeneration occurs; microspores or pollen 
grains die at a late stage of development, and as a consequence of this late death are 
little smaller than surviving pollen grains, and contain collapsed and deep-staining 




triad 
Text-fig. 5. — A-type pollen. The five categories of tetrads. 



cytoplasm. This type of death is sporadic in its occurrence and is apparently unrelated 
to the basic pollen development system of the tribe. Consequently, in scoring tetrad 
type frequencies late death has generally been ignored. 

It could well be that full development of pollen grains at the time of anther 
dehiscence, as judged by their size and the condition and stainability of their 
cytoplasm, is not a consistent indication of their functional viability. Only one of the 
two or three grains in dyads and triads might in fact be functional. Observations 
on pollen germination in nectar (Table 4) deny this possibility. The two grains in 
any dyad each have a germination potential equal to that of single pollen grains in 
monads and are independent of each other in this respect (x^ = 1-01, P = 0-4). It 
seems likely that the same independence would apply to the several pollen grains in 
triads and in full tetrads. 



Tetrad Segregations in A-Type Pollen. 
In A-type pollen the tetrahedral arrangement of the young microspores and the 
distortion caused by the growth of the good grains prevent any possibility of the 
recognition of the planes of the first and second divisions. Analysis of the segregations 
is thereby severely limited and can be made only on the basis of unordered tetrads. 
The shapes of the frequency distributions of the five categories of tetrad in the 
various species, and even in different plants within species, are often very different. 



20 



POLLEN DEVELOPMENT PATTERNS IN THE EPACEIDACEAE, 



as will be seen from an inspection of the data in Tables 5, 6 and 7 and from the 
photographs in Plate ii. In respect to segregation characteristics, each species requires 
individual study and analysis. Such analysis vfiW not be attempted here. 

In Tables 5 and 6 a selection has been made to illustrate the diversity of 
segregation behaviour in different plants of the two species of Astroloma, and Table 7 
presents data for most of the other species which were listed in Table 3. The 




Text-flg. 6. — Pollen development in Astroloma 21 mi folium, x ca. 1150. 



selection of data in each table has been made to emphasize the fact that there is 
an almost complete range, from distributions with a modal class of monads (e.g. 
A. pinifoUum 58/1005, Table 5) to others with a modal class of full tetrads (e.g. 
Leucopogon melaleucoides and Bracliyloma scortecliinii. Table 7). A-type pollen 
segregations thus connect the type of behaviour previously described in Leucopogon 
virgatus with the regular full tetrad T-type pollen. There is in fact a continuous 
series of pollen types within the Styphelieae, from the extreme S-type with nuclear 
migration characteristic of Styplielia itself, through the S'-type to S'-type with break- 
down, leading to A-type variable tetrads and finally to regular full tetrads. 



BY S. SMITH-WHITE. 



21 



Dioecism and Gynodioecism in the Styphelieae. 
Before I undertake a general discussion of the problem of pollen development 
which I have outlined, I want to describe another feature of the Styphelieae which 
may appear to be quite unrelated. This is the occurrence of dioecism and of 
gynodioecism in several species. 

Table 4. 

PoUen (/ermination in Astroloma pinifolium Benth. 

A. In Monads. 



Germinated. 


Not 
Germinated. 


Total. 


Percentage 
Germinated. 


181 


150 


331 


54-8 







B. In Dyads. 






Both P.G.'s 
Germinated. 


One P.G. 

only 
Germinated. 


Neither 

P.G.'s 

Germinated. 


Total 
Dyads. 


Percentage 

P.G. 
Germination . 


41 


59 


34 


134 


.52-6 


Expect 
37-08 


ed on basis of i 
66-81 


ndependence. (p + q)^. 
33-11 






X^ = l-91. 


P = 0-4 







Table 5. 
PoUen Tetrad Segreoation Data. Astroloma pinifolium Benth. 
A. Bast Coast Districts. 



Plant. 



Os. 



Tetrad Type Frequencies Percentage. 



Is. 



2s. 



3s. 



4s. 



Percentage 

PoUen 
Fertility. 



A. East Coast Districts. 



E51/1 




4-0 


30-8 


48-7 


15-3 


1-2 


2336 


44-7 


•W54/9 




16-2 


39-7 


.32-8 


10-5 


0-8 


3464 


35-0 


A58/1 




20-8 


35-1 


34-9 


8-6 


0-6 


11322 


33-2 


055/1 




20-0 


39-8 


30-3 


9-1 


0-8 


1721 


32-7 


W54/5 




26-5 


43-3 


25-0 


4-7 


0-3 


3603 


27-2 


M53/3 




40-0 


37-3 


18-7 


3-7 


0-3 


8022 


21-5 


W54/6 




60-1 


31-5 


7-5 


0-85 


0-05 


5878 


12-3 



B. Grampians District, Victoria. 



58/1008 .. 


6-60 


16-57 


75-79 


0-33 


0-71 


2408 


43-0 


58/1001 . . 


4-31 


20-82 


74-16 


0-05 


0-66 


1974 


43-0 


58/1005 .. 


12-29 


71-49 


16-14 


0-04 


0-04 


2571 


26-0 


58/1007 . . 


30-28 


59-46 


10-26 


0-00 


0-00 


2758 


20-0 


58/1016 .. 


16-61 


71-10 


12-25 


0-04 


0-00 


2776 


23-9 



Localities : A., Audley. B., Evans Head. 0., Oatley. W., Warrah. M., Malabar. 

The majority of flowering plants are hermaphrodite. Dioecism is comparatively 
rare, but perhaps 75% of the families, and nearly all the larger families, include some 
dioecic species (Yampolsky, 1922; Lewis, 1942). Whether the primitive angiosperm 
was unisexual or bisexual may never be known with any certainty, but in many 
families there can be little doubt that dioecism is a secondary and derivative condition. 

Dioecism makes outcrossing obligatory. It ensures a high rate of genetic 
recombination and allows the maintenance of genetic heterogeneity and adaptability. 



22 



POLLEN DEVELOPMENT PATTERNS IN THE EPACRIDACEAE, 



Taele 6. 
Tetrad Segreijation Data. Astroloma coiiostephioides F. MuHl. 




A. Adelaide District. 



65/107 


3-8 


300 


64-2 


0-9 


11 


1872 


41-4 


55/115 


19-8 


20-9 


47-9 


5-2 


6-2 


3185 


39-3 


65/116 


20-5 


55-1 


22-3 


1-7 


0-4 


2405 


26-6 


55/104 


46-5 


51-6 - 


1-9 


0-0 


00 


3200 


139 


55/101 


61-0 


37-2 


1-8 


0-0 


0-0 


2479 


10-2 



B. Grampians District. 



68/214 


8-5 


13-2 


77-1 


0-1 


11 


1621 


43 


58/201 


0-5 


46-8 


35-6 


11-9 


0-6 


1281 


39-0 


68/401 


8-8 


30-8 


59 • 


0-7 


0-7 


2843 


38-4 


58/210 


15-4 


58-6 


25-5 


0-0 


0'5 


2182 


27-9 


58/212 


14-0 


72-0 


13-9 


0-0 


0-1 


1637 


25 1 


58/101 


18-1 


66-8 


15-0 


0-0 


0-1 


2102 


24-2 



Tahle 7. 
Examples of Tetrad Seuregation in Other Species. 







Tetrad-type Proportion? 


Percentage 






Percentage 


Species and Plant. 












N. 


Pollen 














Os. 


Is. 


2s. 


3s. 


4s. 




Fertility. 


Acrotriche 
















cordata R.Br. 
















58/1 


0-94 


] ■ 20 


31 - 89 


33 - 76 


23-21 


853 


67-0 


depressa R.Br. 
















58/1 


23-35 


54 - 39 


21 -22 


1-04 


- 00 


3554 


25-0 


fasciculiflora Bentli. 
















54/6 


4-08 


32 ■ 32 


52-74 


10-67 


0-18 


1640 


42-6 


54/5 


7-84 


40-67 


45 - 79 


6-06 


0-00 


2326 


37-6 


54/8 


18-24 


43 ■ 68 


32-31 


5-61 


0-16 


1891 


31-4 


54/12 


90-12 


7-39 


2-44 


0-05 


0-00 


4008 


3-1 


JSrac/njloma 
















ericoides Sond. 
















58/0 


3-96 


7-74 


82-64 


1-23 


4-43 


1060 


48-6 


58/1 


1-62 


6-94 


90-71 


0-06 


0-67 


1786 


47-8 


58/2 


22-18 


57-54 


20-04 


0-12 


0-12 


1677 


24-6 


preissii Sond. 
















50/49 


3-07 


31 • 68 


45 ■ 00 


20-15 


0-09 


1171 


45-6 


52/16 


2-29 


23-46 


53 - 66 


20-52 


0-07 


1530 


48-2 


52/35 


19-90 


48-06 


27-75 


4-29 


0-00 


955 


29-1 


scortechinii F.M. 
















51/1 . . . . • , . . 


0-43 


1-44 


5-33 


21-61 


71-18 


694 


90-4 


Leucopoi/on 
















melaJevcoides A. Cunn. 
















T58/12 


4-82 


21-72 


34-80 


29 - 59 


9-07 


3342 


54-1 


/20 


3-20 


14-84 


32-24 


35-03 


14-69 


3942 


55-7 


/15 


3-95 


18-73 


27 - 57 


30-83 


18-92 


2024 


60-9 



It is the predominant feature of tlie sexual systems in animals, but in plants the 
same end is achieved more often by other means, and particularly by self-incompatibility, 
which are perhaps more suited to the circumstances of non-mobility and the mechanisms 
of pollen transference (Mather, 1940; Westergaard, 1958). 

Gynodioecism is a system v^^here a population or a species contains individuals 
which are hermaphrodite and others which are male sterile, i.e. functionally female. 



BY S. SMITH-WHITE. 23 

Sporadic male sterility is known in many normally hermaphrodite plants (Lewis, 
1942). Established gynodioecism as a feature of the breeding system of a species has 
been studied in about a dozen species of flowering plants and is known to be 
particularly frequent in the Labiateae. Frankel (1940) suggested that such systems 
may be more frequent in other families than is indicated by actual records. 

Normal dioecism is maintained in natural populations by the segregation of a 
pair of allelic sex genes, linked blocks of genes, or sex chromosomes. Lewis (1941) 
has shown that gynodioecic systems cannot be easily maintained on the same basis — 
stability can only be achieved where the male-sterile plants are more than twice as 
seed-fertile as the herm^aphrodite ones, and the frequency of male-sterile individuals 
must be less than 50% of the population. Chromosomal (or genie) gynodioecism is 
frequent in experimental stocks of cultivated plants, but is rare in natural populations 
(cf., however, Lewis and Crowe, 1956). On the other hand, cytoplasmic systems of 
gynodioecism are more frequent in wild species. In such systems, maternally inherited 
cytoplasm determines whether an individual will be hermaphrodite or male-sterile. 
The method permits a delicate control of outcrossing without being dependent upon an 
extreme advantage in seed set of male-sterile plants. 

In the Ericales the normal condition is monoecism with perfect bisexual flowers. 
Dioecism is rare and must be derived. However, Bentham (1869) observed that some 
species of the Styphelieae show "partial dioecism". Apart from the normal condition 
of hermaphroditism, three rather different modifications of the sexual system occur 
within the tribe. These are: (i) complete dioecism, (ii) gynodioecism comprising 
hermaphrodite and male sterile plants, and (iii) androgynodioecism, v/here hermaphro- 
dites, male-sterile (female) plants, female-sterile (male) plants and male-sterile/female- 
sterile (neuter) plants occur in the population. 

In Monotoca eUiptica R.Br, at Woy Woy, N.S.W. (Smith-White, 1955c) male- 
fertile plants are completely seed-sterile and dioecism is complete in the functional 
sense although poorly developed pistils are present in the flowers of male plants and 
antherless staminodes are present in the females. Bentham (I.e.) has stated that in 
coastal populations of this species dioecism is incomplete, but in the Woy Woy 
population male-fertile plants are completely seed-sterile. The frequencies of male 
and female plants in the population are equal — one count gave 138 male and 146 
female plants — and a simple sex gene or sex chromosome segregation is indicated. 

A similar functionally dioecious system occurs in Monotoca scoparia R.Br, at 
La Perouse, and possibly also in Leuco'pogon hookeri Sond. at Kosciusko. Godley (1957) 
has reported dioecism in Cyathodes spp. and in Leucopogon fasciculatus in New Zealand. 
Cyathocles divaricata Hook, and C. parviflora R.Br, in Tasmania also include male 
sterile plants (Smith-White, 1955c) but it is not known whether dioecism is complete 
in these two species. A system of gynodioecism including female and hermaphrodite 
plants probably occurs in Leucopogon melaleucoides A. Cunn. at Torrington, N.S.W. 
(McCusker, unpublished data). In Leucopogon hookeri Sond. at Barrington, however, 
male, female and hermaphrodite plants exist. 

The sexual system in Lissanthe montana R.Br, at Kosciusko is still more complex, 
the population including hermaphrodites, males, females and neuters (Text-fig. 7). 
There is, however, a complete intergradation of sex types, since male plants show 
variation from "strong" to "weak" in respect to the size and pollen content of anthers, 
and female plants vary in pistil development and fertility. Hermaphrodite plants 
usually have less well-developed pistils than females and smaller anthers than males. 

In field counts of the frequencies of male-fertile and male-sterile plants the 
following results were obtained. 

Male-fertile. Male-sterile. 

5 + c? S + (8) 

At Kosciusko Summit 37 63 

Roadside, 4J miles below summit 58 42 

Roadside, 6 miles below summit 48 32 

Charlotte's Pass 56 44 

Spencer's Creek 58 42 



24 



POLLEN DEVELOPMENT PATTERNS IN THE EPACKIDACEAE, 



With the exception of the Summit counts, the frequencies are consistent and 
indicate a preponderance of male and hermaphrodite plants. 

Sexual differentiation in Monotoca scoparia connects the complete dioecism of 
M. elliptica with the sex-strength system of Lissanthe montana, since male plants in 
some populations of M. scoparia show a limited seed set. There can be little doubt that 
the systems of sexual polymorphism found in the several species of the Styphelieae 
are related. 




Text-fig. 7. — Representative pistils, ovules, and anthers from male, female, hermaphrodite 
and neuter plants of lAssantJie montana. x 20. The two female-sterile plants represented 
differ slightly but consistently in the size of their aborted ovules. 

The occurrence of complete dioecism in some species or in some populations, and 
of male and neuter plants in others, precludes the operation of a cytoplasmic system 
of sex determination. A cytoplasmic system of gyuodioecism does not and cannot lead 
to complete dioecism since with the latter the cytoplasm of the whole species must 
be of the maternal, female type. A complex genetic segregational system of 
determination must be present in Lissanthe montana. Such a system may have become 
progressively simplified and stabilized to give complete dioecism in Monotoca. 

Crane and Lawrence (1931) have described an experimentally produced sexual 
system in Rubus iclaeus which depends on segregation at two independent gene loci. 
Males are ffMm or ffMM, and females are Ffmm or FFmm. Crosses between females 
Ffmm and males ffMm yield both hermaphrodite FfMm and neuter ffmm individuals. 
A similar system is known to occur in Vitis species (Negrul, 1936; Oberle, 1938). 
Another two-factor model is suggested for the Labiate Origanum vulgare by Lewis and 
Crowe (1956). These cases provide models for the interpretation of dioecism and 
gynodioecism in the Styphelieae, with the proviso that the actual system must be a 
good deal more complicated in Lissanthe. 

General Discussion. 
1. The Evolutionary Relationships of the Pollen Types. 
The maturation of pollen in tetrads is not uncommon in the Angiosperms. Some- 
times it may be characteristic only of individual species or genera, as in Leschenaultia 
(Goodeniaceae, Martin and Peacock, 1959, in MSS.) and in Acacia (Mimosaceae, 



BY S. SMITH-WHITE. 25 

Newman, 1934), or it may be characteristic of whole families such as the Winteraceae, 
the Droseraceae, and the Juncaceae. A survey of its occurrence (Wodehouse, 1936; 
Brdtman, 1952) establishes the fact that it must have had many independent origins. 
Indeed the establishment of the tetrad form from the more usual single condition 
requires only the persistent cohesion of the post-meiotic microspores, and in genera 
which normally produce single pollen grains unusual conditions may lead to this 
cohesion. Thus La Cour (1949) has shown that heat shock may have such an effect in 
Scilla. Levan (1942) has described a gene in Petunia which conditions the maturation 
of tetrad pollen. 

Since tetrad pollen is generally characteristic of the Order Ericales, excepting 
only the monogeneric Clethraceae, it is a possible but by no means necessary assumption 
that it had a monophyletic origin approximately contemporaneous with the origin of 
the Order. Had there been separate origins of the character within the several 
families or tribes it might be expected that several taxa within the Order would have 
retained the true single pollen condition, and this does not appear to be the case. 

Any assumption of multiple or polyphyletic origins of monad type pollen in the 
Styphelieae is not permissible. The type is not found elsewhere among the Angiosperms. 
Only in the Cyperaceae is there any parallel behaviour (Juel, 1900: Stout, 1912; 
Piech, 1928; Tanaka, 1940, 1941), and the details of pollen development in the two 
families are quite different. Within the Styphelieae — and also within the Cyperaceae — 
the monad form must have had a single origin from a prior tetrad form, and this 
origin must have been contemporary with the origin of the tribe itself. The descriptions 
which have been presented earlier show that the monad pattern of development must 
be determined by a quantitative cytoplasmic system which must itself be under 
ultimate nuclear control. The whole system must be complex, and it would seem 
more likely that it could break down on several occasions than that it should have 
had several independent origins. 

Since both full tetrad and monad pollen types occur within the tribe, and since 
the latter must be derived from an earlier tetrad condition, it could be supposed that 
the tetrad type, where it occurs within the tribe, is actually primitive. Alternatively, 
the hypothesis may be suggested that the tetrad type within the tribe results from 
the loss of the conditions necessary for monad development. The first view cannot be 
maintained. Table 8 summarizes the three-way relationship between genera, basic 
chromosome number and pollen types. The genera were established by Robert Brown, 
and although later botanists have not always recognized them as having generic status 
(Mueller, 1868; Drude, 1898; Maiden, 1916) they have never denied their reality as 
natural groups and have merely reduced them to sectional rank within a single 
genus Styphelia. 

It has been inferred (Smith-White, 1955a) that the basic numbers in the Styphelieae 
are x = 4 and x = 6, and that the numbers 7, 9 and 10 are derived by polyploidy and 
secondary change. This view is necessary in Astroloma particularly, where the basic 
haploid number is 4 in fourteen species all with monad pollen, and where two species 
have n = 7 and segregating tetrad pollen. 

A consideration of the information in Table 8 shows that if the full tetrad T-type 
pollen is truly primitive within the tribe it is necessary to assume that the S-type 
has originated on several occasions or that the recognized genera have no phyletic 
meaning. It would also mean that the haploid numbers 7, 9 and 10 must be more 
primitive and that the numbers 4 and 6 must have originated several times, and 
always in association with an origin of the monad pollen type. Such a view is 
untenable. 

Accepting the view that the origin of the monad type was contemporaneous with 
the origin of the tribe, and that the primitive basic number was x = 6, the S-type 
pattern was not upset by the origin of the 4-chromosome genom. The full tetrad 
T-type pollen in species of Acrotriclie, Lissanthe and Leiicopogon must then represent 
separate and independent breakdowns of the monad pattern, and it is significant that 



26 



POLLEN DEVELOPMENT PATTERNS IN THE EPACEIDACEAE, 



such breakdowns are always associated with the origin of secondary basic numbers. 
Also, the S'-type of development, the partial breakdown in L. virgatus, and the 
segregating tetrads in twelve species belonging to four genera must represent stages 
in a transition back from monad to full tetrad pollen. This reversion does not do 
violence to the general concept of the irreversibility of evolution. 

It is known that tetrad segregation in Astroloma jnnifolmm and in Astroloma 
conostephioides is associated with abnormal chromosomal conditions and behaviour 
during meiosis, and these abnormalities are in part responsible for the observed 

Table 8. 
Genera, Pollen Types and Chromosome Numbers In the Styphelieae. 



^~--^_CHROMOSOME 


4 


8 


16 


6 


12 


24 


lO 


II 


7 


14 


9 


Un- 
known 


Styphelia 


¥' 
























Astroloma 


M' 


1 


1 












M 








Conostephium 




M 






















Melichrus 




2 






















Leucopogon 

jS PIcuranlhus 


7 


1 




H 






m' 

1 












Leucopogon 

J^ PerojoQ 








M 

4 


M 

2 


M 

1 


MT 

1 1 


^ 










Monotoca 










H° 
















Cyathodes 










M 

2 




2 








T 

1 




Leucopogon 

jf H«tcranthGsis 




















tv: 




y» 


Lissanthe 


















T 

2 


1 




V 

1 


Pentachondra 




















T 






Brachyloma 


















V 

1 




MT 

2 1 


V 

2 


Acrotriche 






















\\ 


V3 


Trochocarpa 














T 

1 












TOTALS 

NUMBERS OF SPP 


28 


6 


1 


M 

9 


M 

5 


M 

1 


MT 

2 1 


M' 

9 


TV 

2 3 


TV 

3 1 


MT 


V 

7 



M MONAD TYPE V VARIABLE POLLEN TETRADS 

M MONAD TYPE Microspores inilmlly unequal I REGULAR FULL TETRADS 

M MONAD TYPE Microspores iniUolly equal D SPECIES WITH OIOECISM or PARTIAL OIOECISM. 

SUBSCRIPTS- Nos o< Species ^ POLLEN TYPE UNKNOWN 



segregations. In Lissanthe strigosa only a single plant amongst more than fifty is 
known to show segregation and in this species it is necessary to assume that the 
segregation is derived directly from the full tetrad condition. If, however, the monad 
type is t.he primitive one within the tribe, unusual cytoplasmic conditions must be 
associated with segregation, and in some of the species, particularly in Astroloma and 
Brachyloma, segregation must be derived directly from the monad type. It is probable 
that the behaviour, even when not directly derived from the monad type, is dependent 
upon a preadapting evolutionary history involving monad pollen, cytoplasmic differen- 
tiation or polarity, and possibly the cytoplasmic control of chromosome segregation. 
This view is further supported by the high frequency of the occurrence of tetrad 
segregation within the tribe compared with its low frequency in other families with 



BY S. SMITH-WHITE. 27 

tetrad pollen, and by the quantitative intergradation which connects the extreme 
monad and the full tetrad types. 

2. A Deductive Hypothesis. 
The treatment of the problem of pollen development in the Styphelieae so far has 
been descriptive and inductive. The main conclusions which have been reached may 
be enumerated. 

Firstly, all the patterns of pollen development iu the tribe are related in causation. 
They constitute parts of one large problem. One general hypothesis must be capable 
of explaining and relating all the modifications of this causal system, and an attempt 
to formulate such an hypothesis is worth while. 

Secondly, the monad pattern of pollen development has had a single origin, roughly 
contemporaneous with the origin of the tribe itself, and may be regarded as the 
primitive condition within the tribe. Although the monad type must have originated 
from an earlier tetrad condition, full tetrad pollen, where it occurs within the tribe, 
is derived by the breakdown of the monad type. Tetrad segregation must also be 
related to the monad type. 

Thirdly, from a consideration of the implications of the Styphelia type of monad 
development, it is inferred that the control of the monad system is primarily 
cytoplasmic and there must be a cytoplasmi(; polarity within the mother cell. 
Cytoplasmic polarity must mean that certain vital constituents attain a gradient 
distribution within the general "body" cytoplasm of the cell. Chromosomal homo- 
zygosity may be assumed as an initial condition, and differential nuclear migration is 
determined by position. 

Fourthly, the permanent triploid L. juniperinus demonstrates that unbalanced 
univalent chromosomes are sensitive to this gradient and control of their segregation 
follows as an immediate consequence. Any control of the segregation of homozygous 
paired chromosomes is impossible and meaningless, but such control may extend to 
structurally heterozygous bivalents provided this heterozygosity involves sensitive 
chromosome segments. Support for this suggestion comes from evidence discussed 
earlier (page 13), that preferential segregation is often a function of heterochromatin. 

Fifthly, the triploid species also shows that a similar control operates in the 
embryo sac mother cells, but under the influence of a different tissue environment. 
The conditions necessary for complementary gametic selection are present and confer 
preadaptation to permanent hybridity. 

Finally, cytoplasmic differentiation must be under genotypic control. It is 
quantitative and complex. Change in genetic balance consequent upon change in the 
basic genom predisposes to the loss of polarity in the mother cell. 

The Initial Operation of the System. 

These conclusions may be taken to constitute a fundamental system and an 
attempt will be made to deduce its evolutionary consequences. Text-figure 8 illustrates 
the immediate operation of the system in pollen development in a homozygous diploid 
with four pairs of chromosomes. Polarity causes a gradient distribution of certain 
cytoplasmic constituents (stippling) and also of non-chromosomal nuclear materials. 
Nuclear migration is one result. Microspore survival is determined by the positions of 
the nuclei in the mother cell; that microspore survives which receives an adequate 
supply of both the polarized materials and of the "body" cytoplasm, and this would 
obviously be the isolated nucleus at the "negative" end of the polarity, since, getting 
the major share of the body cytoplasm, it would also get sufficient of the polarized 
constituents. 

The operation of the same system in the ovule is illustrated in Text-figure 9. 
Intracellular conditions in the P.M.C. and iu the E.S.M.C. are the same, but the 



28 



POLLEN DEVELOPMENT PATTERNS IN THE EPAOEIDACEAE, 



closely enveloping ovule tissue maintains a linear arrangement of the second division 
spindles and of the megaspores, and precludes nuclear migration. An ovule tissue 
gradient, not present in the anther, permits the development only of the micropylar 
megaspore. 

Conditions of megaspore competition are thus different from those of microspore 
competition. The intracellular E.S.M.C. polarity may favour either the micropylar 
or the chalazal ends. In the former case (Text-flg. 9, left) the micropylar megaspore 
receives a major share of the polarized cytoplasmic components and an equal quarter- 
share of the body cytoplasm. A functional em-bryo sac is produced. Where, however, 
the intracellular polarity favours the chalazal end and operates against the tissue 
gradient, the micropylar megaspore gets insufficient polarized constituents and aborts; 
the micropylar megaspore commences to develop, but fails at an early age, and no 
functional embryo sac is produced (Text-fig. 9, right). The direction of the intracellular 




Text-flg. 8. — The control of pollen developing in a diploid species. A cytoplasmic polarity 
(stippling) , determines nuclear migration and survival. Cf. text. 



mother cell gradient is apparently independent of the ovule tissue gradient, since 
species with monad pollen show a maximum of 50% seed fertility. 

The hypothesis set up is admittedly superficial, since the nature of the gradient 
cannot be suggested. It is, basically, an hypothesis of the interaction of two different 
kinds of competition, balanced against each other differently in the pollen and embryo 
sac developmental sequences. Functional pollen grains are developed only from the 
"negative" end of the cytoplasmic axis, and functional embryo sacs come only from 
the "positive" end of the same axis. If unbalanced chromosomes are sensitive to cell 
polarity there is preadaptation to the maintenance of permanent hybridity. 

The Operation of the System in the Triploid. 

The operation of this system in the triploid Leucopogon is illustrated in Text- 
figures 10 and 11, and these figures need very little further explanation. When the 
triploid was first formed, presumably following hybridization, the exclusion of univalents 
from the effective pollen, and their inclusion in the effective embryo sacs would be 
an immediate consequence of the prior conditions of pollen and embryo sac development. 



BY S. SMITH-WHITE. 



29 



This initial control probably is not the final one. The univalents would be subject 
to a predominantly maternal inheritance and, since lethal mutations are relatively 
frequent, they could in the course of time accumulate pollen lethals preventing the 
survival of the exceptional grains which might contain them. Selection would also 
favour the accumulation in them of genes favourable to embryo sac development, and 
there would then be no retreat from permanent hybridity. 

It is possible that a similar system is responsible for the maintenance of univalents 
in the Rosa canina complex. In these roses there is no polarity in pollen development, 
and univalents are eliminated in the pollen line only by the intolerance of genetic 
unbalance. In the embryo sacs, however, conditions must be very similar to those in 





ch. 



m^ 



Text-fig. 9. — The control of embryo sac development in a diploid species. There is inter- 

accion between an ovule tissue gradient favouring the micropylar megaspore (bottom arrows) 

and a cytoplasmic polarity, (stippling). The latter may favour either end of the megaspore- 
telrad. Cf. text. 

L. juniperinus and an initial tendency for the univalents to be inherited in the 
maternal line would then be strengthened by selection both of genes necessary for 
embryo sac development and of more reliability in polarized segregation, as suggested 
by Darlington (1939). 



The Oric/in of Permanent Hyhrtdity in Diploid Species. 

If, as has been inferred, cytoplasmic control of segregation can extend to 

structurally heterozygous bivalents, it would be possible for permanent structural 

hybridity to be established in diploid species possessing monad pollen. It will be 



30 



POLLEN DEVELOPMENT PATTERNS IN THE EPACRIDACEAE, 



asssunied that structural hybridity may arise adventitiously in any species, but that 
in sexual species with a normal genetic system it will usually be eliminated. 

One necessary condition, for permanence, is some degree of chiasma localization, 
but this is not a serious restriction, since chiasmata are rarely or never strictly 




Text-fig. 10. — The control of pollen development in the triploid heucopogon juniperinus. 
Univalents are "gradient-positive" and are excluded from the functional pollen. Cf. text. 





ch. 



m. 



Text-fig. 11. — The control of embryo-sac development in the triploid Ijeucopogon juniperinus. 
Tile univalents, being "gradient-positive", are included in the functional embryo sacs. The 
univalents of the triploid are strictly maternal in inheritance. Cf. text. 



BY S. SMITH-WHITE. 



31 



terminal or strictly adjacent to the centromeres. Text-figure 12 illustrates a bivalent 
in which there is one chiasma, with the proximal and distal regions excluded from 
the chiasma region. The bivalent can be extended to a second chiasma region without 
affecting the subsequent argument provided the first and second chiasma regions do 
not overlap, i.e. provided there is a degree of chiasma localization. An adventitious 
structural change, adjacent to the centromere, would cause an orientation of the bivalent 
at first metaphase, with the more sensitive segment directed towards the positive end 
of the cell polarity. As with the univalents in the triploid, this controlled segregation 
would favour the exclusion of the positive reacting chromatin from the pollen and 
its inclusion in the embryo sacs, and the negative responding chromatin would be 
favoured in the pollen. Similarly, an adventitious structural hybridity distal to the 
chiasma region, and perhaps with neocentric activity or acting indirectly through the 
centromere (Darlington, 1956), could cause a non-random orientation of the chromatids 
at the second division. It is not necessary that this control should be near-perfect in 
the beginning, but only that it should be substantial, since selection would then favour 
its greater stabilization. Adventitious structural hybridity would be maintained by the 
complementary gametic selection imposed by the cytoplasmic system. 




Text-fig. 12. — Diagram of a bivalent with a single chiasma and some localization, 
proximal and distal segments lie outside the chiasma region. Cf. text. 



The 



With this cytoplasmic system operative, a reinforcement with a genetic mechanism 
would be almost inevitable. Bennett (1956) has demonstrated that even in species 
with more normal genetic systems there is an appreciable prospect of the establishment 
of balanced lethal mechanisms within non-crossover segments. In the system suggested 
here there would be no selection against pollen-lethal mutations in the maternally 
inherited gradient-positive segments nor against embryo sac lethals in the paternal 
gradient-negative segments. This would lead to the reinforcement of the cytoplasmic 
control of the monad pattern of development by a more rigid system involving chromo- 
somal segregation and permanent balanced gametic lethals. Referring again to Text- 
figure 12, if A and a represent "positive" and "negative" proximal segments, and if 
B and b represent similar segments distal to the chiasma region, we have precisely the 
segregational scheme mentioned and rejected earlier (Text-figure 2b). 

There is an important complication, however. In Styphelia there are four pairs 
of chromosomes, and each is equally liable to the consequences of cytoplasmic polarity. 
In each bivalent there is a single proximal position (i.e. on both sides of the 
centromeres) and two distal positions (one on each chromosome arm) which are 
protected from crossing over and which are therefore potential sites for structural 
hybridity. With maximum complexity the constitution of a diploid Styphelia or 
Astroloma could be: 

SAT UBV WCX YDZ 



sat ubv WCX ydz 
where the dots represent centromeres and the symbols represent "positive" (capitals) 
and "negative" (lower case) segments. This segregational system could reinforce, 



32 POLLEN DEVELOPMENT PATTERNS IN THE EPACRIDACEAE, 

but not replace, cytoplasmic control, since it would break down if subjected to random 
segregation. Further, it could not operate on the taivalents in the caninae roses, since 
there is no polarity and therefore no possible control of bivalent segregation, in the 
pollen mother cells of Rosa. 

The Breakcloivn of the System. 

It has been found that loss of the monad pattern is associated with secondary 
polyploidy and change of genetic balance. Secondary polyploidy does two things. It 
brings a reduplication of a large part of the genom and it causes a loss of polarity 
in the pollen mother cells. In Astroloma, with x = 4, the two species with A-type 
pollen both have n = 7. Assuming the full structural hybridity illustrated above, their 
constitutions would become 

SAT UBV WCX YDZ U' B' V W C X.' Y' D' Z' 



sat u b V WCX y d z u' b' v' w' c' x' y' d' z' 

It is not necessary to assume that genetic differentiation should occur in all seven 
proximal positions and all 14 distal positions. The loss of cell polarity would open the 
system to segregation. Assuming segmental heterozygosity in all positions except those 
in the unrepeated chromosome, and either duplicate or alternate interactions between 
the repeated parts of the genom (i.e. either auto- or alloploidy), pollen tetrad 
segregation would result, with frequencies of the five possible tetrad types: 

Percentage frequencies. 



Interaction 


nullads 


monads 


dyads 


triads 


full tetrads 


Duplicate 
Alternate 


75-73 
97-10 


24-64 
2-83 


2-50 
0-022 


0-127 
0-0002 


0-003 
0-0000007 



Thus, following loss of polarity, the initial pollen fertility would be extremely low 
provided full complexity had been attained. However, selection should tend gradually 
to undo the system previously built up under the protection of polarity, and simpler 
patterns of segregations such as those actually observed would result. Ultimately, the 
whole segregating system might be eliminated, giving reversion to the full tetrad type. 
It would be expected that the interaction relationships of the reduplicated genoms 
would be different in the several independent origins of secondary ploidy within the 
tribe. Also, the selective reduction of the segregating system would vary in each 
case; it would be rapid in the absence of the accumulation of pollen and embryo sac 
lethals on the "negative" and "positive" segments respectively, but slow and perhaps 
almost impossible in the event of such accumulation. The different species would 
show differences in the ease and rate of reversion to the full tetrad condition. 

The Origin of Dioecism and Mixed Sexuality. 

The chromosomal system wherein certain segments are inherited solely through 
the embryo sacs and where alternative segments are solely paternal, has some similarity 
with ordinary sex mechanisms. A very simple system of mixed sexuality can be 
deduced on such a basis. 

In the simplest case, structural hybridity A/a may be assumed in a single segment, 
where A is positive to cell polarity and a is negative. Segment A has accumulated 
deficiencies which act as gametic lethals in the pollen, and similarly a has accumulated 
embro sac deficiency lethals. With secondary ploidy there is a reduplication of these 
segments, giving either the constitutions A/a A/a (autoploidy) or A/a A'/a' (allo- 
ploidy). This is the simplest form of the system already deduced. There is also a 
loss of polarity. The only new assumption necessary is that in individuals homozygous 
for A/A or A'/A' there is anther abortion, and that in individuals a/a or a'/a' there 
is pistil abortion. Three genotypes of embryo sacs, A A', A a', and a A', and three of 



BX S. SMITH-WHITE. 33 

pollen grains, A a', a A', and a a', arc possible, and in the sporophyte there are seven 
genotypes and four phenotypes: 



A A A A A A' 


A a' a A' 
a a' a a' 

\ 1 


A a' a A' 


A a' a A' a a' 


A a' a A' 


Pier- 
F«™^1^ maphrodite 


Male 


Neuter 



There is an interesting feature of this system. Starting with any frequencies of 
the seven possible genotypes, and in the absence of selection, the system will rapidly 
approach a stable equilibrium in which the proportions p and q of A and a, and p' 
and q' of A' and a', are equal, and in which the four phenotypes have the proportions 
0-24 females, 0-35 hermaphrodites, 0-24 males and 0-17 neuters, which may be further 
summed to 0-59 male-fertile and 0-41 male-sterile. These are close to the actual 
proportions observed in Lissanthe montana at Kosciusko. The occurrence of such a 
stable equilibrium, which of course might be modified by other factors of selection, 
would mean that the system could be readily established once the necessary prior 
conditions were provided. In this respect it resembles the relatively simple case where 
selection favours a heterozygote A/a over either homozygote A/A or a/a. 

The system also resembles fairly closely the two-factor mechanisms which have 
been described in Ruius idaeus and in Vitis, referred to earlier in this address. 
However, it is capable of much greater complexity, where heterozygosity might be 
established in similar non-crossover segments in other chromosomes, and it is capable 
of giving a complex sex-strength system which seems to be necessary in Lissanthe 
lyiontana. It is also capable of being modified in the direction of complete dioecism. 

Stjmmaey. 
The most frequent, and perhaps the most characteristic pollen type in the 
Styphelieae is the monad. Monad pollen is derived from the tetrad form by the 
regular failure of three microspores in the young tetrad. In its most extreme form, 
monad development involves nuclear migration in the mother cell following the 
conclusion of meiosis. 

Other pollen types found in the tribe include modified monads where nuclear 
migration is absent, variable pollen tetrads, and regular pollen tetrads. 

An attempt has been made to show that all pollen types in the tribe must be 
related at the level of causation, and that the extreme monad type is the basic one 
within the tribe. Its establishment must have been roughly contemporaneous with 
the origin of the tribe. The other pollen types represent breakdown or loss of the 
monad pattern of development. 

The monad pattern of development necessarily implies the development of cyto- 
plasmic differentiation or polarity within the pollen mother cell. This polarity may 
possibly be related to that which is normal in angiosperm pollen grains. 

Conditions in a permanent triploid Leucopogon demonstrate that this polarity is 
present in the mother cell during the first meiotic division, and also that unbalanced 
chromosomes (univalents) may be sensitive to it. The permanent triploid also 
demonstrates the existence of a mechanism of complementary gametic selection. This 
mechanism must be antecedent to the origin of triploidy and its presence can be 
inferred in diploids. 

Using these inductive conclusions, an evolutionary hypothesis has been presented 
to explain and relate the different pollen types. According to this hypothesis, structural 
hybridity could be maintained in diploid species by a mechanism similar to that 
demonstrated in the triploid. Loss of polarity, accompanying secondary ploidy and 
change of genetic balance would then give rise to the variable tetrad pollen type. 

The hypothesis permits the deduction of a system of mixed sexuality and unstable 
dioecy which is adequate to explain actual conditions in a number of species of the 
tribe. - . . 



34 POLLEN DEVELOPMENT PATTERNS IN THE EPACRIDACEAE, 

The hypothesis is admittedly a speculative one. Its only merit may be that it 
permits a unification of several problems within the group. It is, however, capable of 
experimental testing, and I hope to be able to contribute in this way, in the future, 
either to its establishment or its downfall. 

Ack7iowledgements. 

I am indebted to Miss A. McCusker for observations and data on pollen tetrad 
segregation in Leucopogon melaleucoides, Lissanthe strigosa and Acrotriche divaricata, 
and to Mr. L. D. Williams, of Meningie, S.A., for bringing to my notice the occurrence 
of tetrad segregation in Acrotriche cordata, A. denressa. and Brachyloma ericoides, and 
for supplying material of these species. 

I am grateful to Miss Margaret Woodward for the preparation of many of the 
diagrams. 

References. 
Barber, H. N., 1941. — Chromosome behaviour in Vvularia. J. G-enet.. 42: 22.3-257. 
Bennett. J. H., 1956. — Lethal gene.s in inbred lines. Heredity, 10: 263-270. 
Bentham, G., 1869. — Flora Auslralieiusis. LoveJl, Reeve d Co., London. 
Brough^ p., 1924. — Studies in the Epacridaceae. i. The lile history of Styphelia longifolia. 

Proc. Linn. Soc. N.S.W., 49 : 162-178. 
BosEMARKj N. O., 1954. — On accessory chromosomes in Festuca pratensis. (ii) Inheritance of 

the standard type accessory chromosome. Hereditas, 40 : 425-437. 
Brock, R. D., 1954. — Fertility In Lilium hybrids. Heredity, S : 409-420 
BrumpihlDj R. T.. 1941. — Asymmetrical spindles in the first microspore division of certain 

angiosperms. Anier. J. Bot., 28: 713-722. 
Catcheside, D. G., 1944. — Polarised segTegation in an ascomycete. Ann. Bot., N.S., 8: 119-130. 

, 1950. — The B chromosomes of Parthenium argentaturii. Genetica Iberica, 2 : 139-148. 

Crane, M. B., and Lawrence, W. J. C, 1931. — Inheritance of sex, colour and hairyness in 

the Raspberry Rubus idaeus. J. Genet., 24 : 243-255. 
Crousb, H. C, 1943. — Translocations in Sciara : Their bearing on chromosome behaviour and 

sex determination. Univ. Missouri Agr. Exp. Stn. Res. Bull., 379 : 1-75. 
Darlington, C. D., 1937. — Recent advances In Cytology, 2nd Edition. Churchill, London. 
, 1939. — The evolution of genetic systems, 1st Edition. Cambridge University Press, 

London. 
, 1940. — The genetical and mechanical properties of the sex chromosomes. (v) Cimex 

and the Heteroptera. J. Genet., 39: 101-137. 
, 1956. — Messages and movements in the cell. Conference on chromosomes. 



Wageningen. 
, and La Cour, L. F., 1941. — The genetics of embryo sac development. Ann. Bot., 

N.S., 5: 547-562. 

, and Mather, K., 1949. — The elements of genetics. Allen and Unwin, London. 

Db Almeida, J. L. F., and Sampata, T. M., 1950. — Sobre a differenciacao nuclear nos micro- 
spores de L^izula purpurea Link. Bol. Soc. Broteriana, 24 : 323-332. 
Drude. O., 1891. — The Epacridaceae. In Engler and Prantl, Die Naturlichen Pfianzenfamilien, 

4 (1) : 76, 78. W. Engelman, Leipzig. 
Erdtman, G., 1952. — Pollen morphology and plant taxonomy. Chronica Botanica Co., Waltham, 

Mass., U.S.A. 
Frankel, O. H., 1940. — Studies in Hebe. (ii) The significance of male sterility in the genetic 

system. J. Genet., 40: 171-184. 
GbitleR;, L., 1935. — Beobachtungen iiber die erste Teilung im Pollenkorn der Angiospermen. 

Planta, 24 ; 361-386. 
GoDLET, E. J., 1957. — Unisexual flowers in the Ericales. Nature, 180 : 284-285. 
GuSTATSON, A., 1944. — The constitution of the Rosa canina complex. Hereditas, 30 : 405-428. 
HOLDEN, J. W. H., and' Mota, M., 1956. — Non-synchronised meiosis in binucleate pollen mother 

cells of an Avena hybrid. Heredity, 10 : 109-117. 
Hurst, C. C, 1931. — Embryo sac formation in diploid and polyploid species of Roseae. Proc. 

R.S.(B), 109 : 126-128. 
KiHARA, H., and Hirayoshi, I., 1932. — Die Geschlechtschromosomen von Humulus japonicus 

Sieb. et Zucc. 8th Congr. Jap. Assoc. Adv. Sci., 1932: 363-368. Cited from Darlington, 1937. 
KiharAj H., and Ono, T., 1925. — The sex chromosomes of Rumex Acetosa. Z.I.A.V., 39 : 1-7. 

Cited from Darlingto.'-,, 1937. 
La Cour. L. F., 1944. — Mitosis and ceil differentiation in the blood. P.R.S. Edin., B, 62: 73-85. 

, 1949. — Nuclear differentiation in the pollen grain. Heredity, 3 : 319-337. 

Lbvan, a, 1942. — A gene for the remaining in tetrads of ripe pollen in Petunia. Hereditas, 

28 : 
Lbwis, D., 1941. — Male sterility in natural populations of hermaphrodite plants. N. Phyt., 

40: 56-63. 



BY S. SMITH-WHITE. 35 

Lewis., D., 1942. — The evolution of sex in flowering plants. Biol. Rev., 17: 46-67. 

, and Crowe^ L. K., 1955. — The genetics and evolution dioecy. Evolution, 10 : 115-125. 

Maheshwari^ p., 1950. — An introduction to the embryology of the angiosperms. McGraio-Hill, 

N.Y. 
Maiden, J. H., 1916. — A census of the plants of New South Wales. Govt. Printer, Sydney. 
Mather, K., 1940. — Outbreeding and the separation of the sexes. Nature, 145: 484. 
■ , 1948. — The nucleus and cytoplasm in differentiation Growth symposium. S.E.B, 

CambrHclge Univ. Press. 
Metz, C. W., 1938. — Chromosome behaviour, inheritance and sex determination in Sciara. 

Am. Nat., 12 : 485-520. 
Mueller, F., 1868. — Fragmenta Phytographiae Australiae, 6 : 50. Govt. Printer, Melbourne. 
MuNTZiNG, A., 1946. — Cytological studies of extra fragment chromosomes in rye. III. The 

mechanism of non-disjunction at the pollen mitosis. Hereditas, 32 : 97-119. 
Negrul, A. M., 1936. — The genetic basis of grape breeding. Bull. Appi. Bot. Select. Sr., 8 (6) : 

1-149. 
Newman, I. V., 1934. — Studies in the Australian Acacias. IV. The life history of Acacia 

Baileyana F.v.M. Proc. Linn. See. N.S.W., 59: 277-313. 
NoviTFKi, E., 1951. — Non-random disjunction in Drosophila. Genetics, 36 : 267-280. 

, 1954. — The compound X-chromosomes in Drosophila. Genetics, 39 : 127-140. 

Oberlb, G. D., 1938. — A genetic study of variations in floral morphology and function in 

cultivated forms of Vitis. Techn. Bull. N.Y. State Agr. Expr. Stn., No. 250. 
OSTBRGREN, G., 1947. — Hcterochromatic B chromosomes in Anthoxanthiim. Hereditas, 33 : 

261-296. 
RHOA0ES, M. M., 1942. — Preferential segregation in maize. Genetics, 27 : 395-407. 
, 1952. — Preferential segregation in maize. In Heterosis. Ed. J. W. Gowan. Iowa 

State Coll. Press, Ames, Iowa. 
Roman, H., 1947a. — Mitotic non-disjunction in the case of interchanges involving the B-type 

chromosome in maize. Genetics, 32 : 391-409. 

, 1947&. — Directed fertilization in maize. Proc. Nat. Acad. Sci., 34 : 46-52. 

Sandler, L., and Braver, G., 1954. — The meiotic loss of unpaired chromosomes in Drosophila 

melanogaster. Genetics, 39 : 335-377. 
Sax, K., and Hustbd, L., 1936. — Polarity and differentiation in microspore development. Amer. 

J. Bot., 23 : 606-609. 
Selling, O., 1947. — Studies in Hawaiian pollen statistics. Part 2. The pollens of the Hawaiian; 

Phanerogams. Spec. Publ. Bishop Mus., 38, Gothenburg. 
Smith-White, S., 1948a. — A survey of chromosome numbers in the Epacridaceae. Proc. Linn. 

See. N.S.W., 73 : 37-56. 

, 19486. — Polarised segregation in a stable triploid. Heredity, 2: 119-129. 

, 1955a. — Chromosome numbers and pollen types in the Epacridaceae. Aust. J. Bot., 

3 : 48-67. 
, 19551). — The life history and genetic system of Leucopogon juniperinus. Heredity, 

9: 79-91. 
— , 1955c. — Cytological studies of the Australian Flora. Thesis for D.Sc.(Agr.), Uni- 
versity of Sydney. 
Sturtbvant, A. H., 1936. — Preferential segregation in triple-IV females in Drosophila 

melanogaster. Genetics, 21: 44 4-466. 
TACKHOLM, G., 1922. — Cytologische Studien ijber die Gattung Rosa. Acta Hort. Berg., 

7: 97-381. 
Westekgaard, M., 1958. — The mechanism of sex-determination in dioecious flowering plants. 

Adv. in Genetics. Ed. M. Demerec. Acad. Press Inc., N. York. 
White, M. J. D., 1950. — Cytological studies on gall midges (Cecidomyidae). Univ. Texas 

Publ. 5007. 

■ — , 1954. — Animal cytology and evolution. 2nd Ed. Cambridge Univ. Press. 

WODBHOUSE, R. P., 1936. — Pollen grains. McGraw-Hill, N.Y. 
Yampolskt, 1922. — Cited from Lewis, 1942. 

EXPLANATION OF PLATES I-II. 
Plate i. 
Breakdown of monad development in Leucopogon virgatus, Plant R57/10. 

Plate ii. 
Variable Tetrad Pollen. 
1. Astroloma pinifoUum x 150. 2, 3. Astroloma conostephioides x 150. 4. Acrotriche 
fasciculiflora x 300. 5. Leucopogon melaleucoides x 300. 6. Brachylomd ericoides x 300. 



36 



PHYSIOLOGICAL SPECIALIZATION OF MELAMPSORA LINI (PERS.) LeV. 

IN AUSTRALIA. 
By H. B. Keee. 
ii - (Plates iii-iv; three Text-figures.) 

[Read 2 5 til March, 19.59.] 



Synapsis. 
A physiologic race survey of the pathogen Melampsora lini (Pers.) Lev. was carried 
out at Sydney University over the period 1948 to 1953. Several important new races were 
recorded indicating a marked change in the pathogenicity of the pathogen in Australia since 
the first surveys undertalven by Waterhouse and Watson. Particular attention is drawn to 
the difference in host range and' geographic distribution of the Punjab-attacliing and 
non-Punjab-attacking race groups. 



Inteoduction. 

Physiologic specialization in M. lini was first demonstrated by Flor in 1935, although 
others were aware of physiologic differences within the species as early as 1865. By 
1935 Flor had identified fourteen races on nine differential varieties. Since then Flor 
has modified and increased his differential series on which he had differentiated 127 
races by 1945 (Flor, 1945). Other races were detected by workers in Canada, Europe, 
Australia, South America and India (Straib, 1939; Waterhouse and Watson, 1943; 
Vallega, 1944; Prasada, 1948). The differential varieties currently used by Flor are 
monofactorially resistant or immune to the races identified by him, and could effectively 
differentiate over a million races which might be realized by random recombination of 
genes already detected in the pathogen. 

According to McAlpine (1906), M. lini was first identified in Australia in 1889 on 
cultivated flax Linum usitatissimum. It was also common on native flax L. marginale. 
He assumed that it had been introduced from overseas. 

Flax and linseed, especially flax, have been grown on a small scale since the 
-earliest colonial days. Acreages remained very small until the 1939-1945 war. Prior 
to 1939 the yearly sowings of flax averaged less than 2,000 acres. This increased 
to a peak of 61,000 acres in 1944, followed by a progressive decline to acreages of less 
than 5,000 acres between 1950 and 1952. Commercial crops of flax were limited to 
Victoria, South Australia, Western Australia and Tasmania. 

Blue Riga, the main flax variety in Victoria prior to 1937, and Concurrent and 
Liral Crown, the dominant varieties in the southern States during and after the war 
years, were highly susceptible to the majority of the Australian races identified since 
1943. Rust-resistant varieties began to replace them (Thomas and Millington, 1946), 
but the pathogen continued to be a serious hazard to flax crops (Cass Smith and 
Harvey, 1946). 

Linseed acreages also increased during the war years and remained at a moderately 
high level for some time afterwards. Moderate acreages were grown in New South 
Wales and Queensland as well as the southern States as late as 1953, when these studies 
were closed. The first commercial crops of Punjab suffered severely from rust, but 
the linseed growing industry was reestablished with Walsh, which has been generally 
resistant to M. lini in all States since its introduction. 

Following serious outbreaks of rust in the early war years Waterhouse and Watson 
(1941 and 1943) commenced a survey of the pathogen. Flor's series of eleven 
differential varieties and Argentine 705-1 constituted the first set of differential 
varieties. The first survey keyed out a unique race which differed significantly from 



Proceedings of the Linnean Society of New South Wales. 1959, Vol. Ixxxiv, Part 1. 



BY H. B. KERR. 



37 



all overseas races in its avirulence on Bison, a variety susceptible to all these races. 
It was designated race A, and was highly virulent on Punjab. Later studies determined 
five additional races, B to F. Most of the collections of rust were from commercial 
crops, but several were from L. marginale. The latter, with one exception, were 
identified as races A or F, both Punjab-attacking and avirulent on Bison. Watson 
continued the survey beyond 1943 but failed to detect any significant change in the 
rust population up to 1945. A further survey was carried out by Charles in 1947 on 
an extended host series of eighteen varieties including the original twelve. This was 
continued by Kerr in 1948 prior to the project dealt with in this paper. These studies 
determined seven additional races G to 0. In addition, Charles reclassified the original 
six races A to F, adding to their description reactions given by them on six additional 
varieties (Millikan, 1951). 

Experimental Methods and Material. 
The differential series previously used by Charles (1947) was adopted during the 
current studies. To these were added Koto, a variety showing promise as a useful 
immune parent, and a selection of Walsh, the only commercial variety of linseed 
grown in Australia. Owing to impurity Williston Brown was later dropped from the 
series- The differential varieties are listed below. 





Sydney University 




Variety. 


Accession 


Number. 


C.I. Number. 


Buda 


Fx 


1 




WiUiston Golden 




Fx 


2 


25-1 


Akmolinsk 




Fx 


3 


515-1 


J.W.S. . . 






Fx 


4 


708-1 


Abyssinian 






Fx 


5 


701 


Kenya 






Fx 


6 


709-1 


Argentine 






Fx 


7 


70.5-1 


Very Pale Blue 


Crimped 




Fx 


9 


647-1 


Ottawa 770B 






Fx 


10 


355 


Argentine 






Fx 


11 


462 


Bison 






Fx 


13 


.389 


Punjab 






Fx 


14 


— 


Walsh 






Fx 


86 


— ■ 


Morye 






Fx318 


112 


Newland 






Fx 


319 


188 


Bolley Golden 






FX320 


644 


Italia Koma 






Fx 


321 


100.5-1 


Leona 






Fx 


322 


836 


Tammes' Pale Blue 




Fx323 


333-1 


Koto 




rx326 


842 



Collections of rust were received from Queensland, New South Wales, Victoria, 
South Australia and Western Australia. The collections were received during spring 
and early to mid-summer. They were cultured on the unnamed variety F257, which 
was found to be more susceptible to the Australian collections than the previously used 
variety, Concurrent. The uredospore inoculum was collected and stored in small, glass, 
cork-stoppered phials in a refrigerator at 0^ to 2°C. 

The studies were carried out in the glasshouses at Sydney University during the 
cooler months of the year from April to late October. Temperature and light intensity 
and duration fluctuated according to conditions out of doors. This was offset by the 
occasional use of radiators and incandescent lights during winter, when the equipment 
was available. 

The reaction of the differential varieties was read after about 10 to 14 days, 
depending on conditions during incubation. Flor's system of reaction classification 
was adopted as far as possible, but the wide range of reaction induced by fluctuations 



38 



PHYSIOLOGICAL SPECIALIZATION OF MELAMPSORA LINI IN AUSTRALIA, 



in the environment prevented very fine distinctions. Only three types of reaction 
were finally distinguished: immune (including consistently immune or highly 
resistant), resistant (including the wide range of rather variable intermediate reactions 
from resistant to moderately susceptible), susceptible (including highly susceptible 
reactions sometimes depressed to moderate susceptibility by adverse environmental 
conditions). (See Plate iii. ) 

Each collection was tested at least twice, and particular attention was paid to the 
varieties giving reactions rather sensitive to the environment. The varietal screening 
method was used to confirm the occurrence of new races, along with the use of single 
spore cultures. 

Adequate precautions were taken to prevent contamination of the races in the 
glasshouse. Races were cultured where possible on varieties immune to the other races 









. Mount Ga^h^r f^^^^^^^J'" "' 

flortdok. UncMseo^-—-, \ Pre. 




Text-fig. 1. — Maps of linseed and flax growing States indicating- main flax and' linseed 
growing districts, and districts from which Melanipsora lini specimens were received. 



cultured in the same house. They were subcultured no more than twice a year and 
sometimes only once a year. 

The spore dusting inoculation technique used by Flor (1935) and Waterhouse and 
Watson (1941) did not always give satisfactory results during these studies. A more 
laborious method was adopted, but this resulted in consistently heavy infection. A 
drop of water was placed on the crown of young unfolded leaves of each seedling. 
Uredospores were spread on the surface of water in a petri dish and transferred by 
spatula to the seedling tip. At first poor germination of uredospores stored for more 
than two months sometimes necessitated reinoculation. It had been the practice to 
check the percentage germination of the residual spores left in the petri dish. While 
germination was generally good on the seedling tips, it was not uncommon to discover 
that the spores had generally failed to germinate on the water in the petri dish. This 
was attributed to diffusion of substances from the host tissue into the terminal drop 
of water. Boiled aqueous host extract was found to induce consistently maximum 
germination in stored spores sown on the surface of the extract. Experiment showed 
that it was generally sufficient to spread stored spores on the surface of concentrated 



BY H. B. KERR. 39 

host extract before transfer to the terminal drop of water on the seedling to ensure 
good germination. When spores had been stored for long periods it was sometimes 
sufficient to leave the spores on concentrated host extract for several minutes before 
transfer to the seedling. To ensure good germination host extract was used to moisten 
the seedling tips. 

The excised shoot technique was used during most of these studies to conserve 
bench space and ensure a ready supply of seedling tips for inoculation (Kerr, 1951). 

During these studies a technique was developed to preserve rust reactions for 
comparison with results obtained later. Leaves were preserved under Scotch tape on 
semi-absorbent paper, and stored in the dark in a refrigerator at 0° to 2°C. After 
two years there was no appreciable deterioration, the leaves and reaction closely 
resembling freshly collected material. 

Experimental Results. 
The Effect of the Condition of the Shoot on the Rust Reaction. 

The condition of the host during the incubation period played an important role 
in determining the reaction of most varieties. The most susceptible varieties usually 
developed a measure of resistance as the shoots became more fibery with age. 

Bison was fully susceptible to every Australian race, except race 1, and F257 was 
fully susceptible to every race in the early seedling stage. Race 6 was cultured on 
both varieties under identical conditions, on young succulent excised shoots of F257 
which remained succulent during the incubation period, and on older excised shoots 
of Bison which were quite succulent when inoculated, but became rather fibery later. 
Both varieties gave a susceptible reaction after 10 days. After three weeks the Infected 
Bison leaves had withered and yielded very little inoculum. F257 continued to yield 
very heavy loads of uredospores. A quantitative assessment of the degree of active 
infection at this time was given by the number of excised stems of each variety with 
prominent uredosori. Only fourteen of thirty-two excised shoots of Bison had uredosori 
on the stem. Thirty-nine of the forty-one shoots of F257 showed prominent uredosori 
on the stems. Bison had obviously acquired some measure of resistance to race 6. 

The degree of succulence of the inoculated shoot was more important than the 
age of the parent plant in determining reaction. Since succulence was largely 
determined by environmental factors, rather than any inherent characteristic of the 
host, during the first month or so after excision of the shoot, the value of standardizing 
growing conditions at the optimal was apparent. 

The Effect of the Environment on Host Reaction. 

Recognition of the significantly different reactions of each variety in the differential 
series to different races of rust is fundamental to the success of any physiologic race 
survey. Since no two varieties behave alike, each must be studied individually to 
determine the number of significantly different reactions and the extent to which each 
reaction is subject to environmentally induced fiuctuations. 

The lack of temperature and light control facilities at Sydney accentuated the 
importance of this aspect of the work. Variations in the environment often induced 
changes in reaction as great as that considered adequate for the differentiation of races. 
The reaction of several varieties to both races 2 and 13 was suppressed in the direction 
of greater resistance under the very warm incubation conditions in the glasshouse 
during late October and early November. The variety Very Pale Blue Crimped was 
immune to race 2 under these conditions, although it gave a fully susceptible reaction 
to race 2 under the optimal conditions (15° to 20°C.) prevalent earlier in the year. 

Races 1 and 2 were tested in the field at Sydney to determine their capacity to 
survive in the uredospore stage during spring and summer. The former was highly 
virulent on Punjab and moderately virulent on F257 in the glasshouse under optimal 
conditions, and the latter was avirulent on Punjab but highly virulent on F257. Both 
were well established on their susceptible hosts in September. Periodic sowings of 



40 PHYSIOLOGICAL SPECIALIZATION OF MELAMPSORA LINT IN ATJSTBALIA, 

both varieties were made to ensure a constant supply of young infectable plants. As 
the temperature increased during November and December, Punjab continued to be 
heavily infected by race 1. The poor infection of F257 reflected the lack of tolerance 
of race 2 to high summer temperatures and the relative immunity of the variety to 
race 1 under these conditions. 

Reaction Levels. 

After careful consideration of all the results obtained with each accession it was 
decided to recognize no more than three distinctly different reactions in each variety. 
In most cases only two significantly different types of reaction were differentiated. 
Some varieties may have had more significantly different reactions than the number 
finally attributed to them. But in view of the wide variations in light intensity and 
temperature during tests and between different tests of each accession, it was considered 
unwise to postulate a greater number than that finally decided upon. 

It is proposed to designate significantly different reactions within a variety 
"reaction levels". There seems to be no conclusive evidence to show whether the 
reaction of a variety to all possible races of a given pathogen ranges by almost 
imperceptible stages from complete immunity to complete susceptibility, or whether 
there is only a limited number of different reactions. If the former is the case the 
term reaction level has little meaning, the number of levels determinable being limited 
only by the degree of environmental control and accuracy of observation. The results 
obtained during this and most other rust surveys, and Flor's genetic investigations of 
virulence and avirulence in M. lini seem, however, to support the thesis that there is 
only a limited number of reaction levels. This assumption underlies the use of the 
term reaction levels. 

In the final analysis races should be differentiated from each other and described 
in terms of these reaction levels. A double descriptive system would seem to be 
necessary. The rust reaction of a variety to any race is the product of the interaction 
of a gene or genes in the pathogen and corresponding genes in the host, modified by 
the environment, and possibly also by modifying genes in the race and host. No two 
varieties are likely to have the same range of reactions. The resistant reaction of 
Walsh was characteristic of that variety and differed noticeably from the resistant 
reaction of Akmolinsk to the Australian races. The initial descriptive terminology 
should be sufficiently comprehensive to define these differences. 

However, fine differences in the resistant reaction of varieties are irrelevant in 
the final race classification. These differences are no more relevant than those induced 
by the environment. Final decisions on race status must necessarily be intravarietal. 
Two races can be dift'erentiated only if the reaction of one differs significantly from the 
reaction of the other on the same variety under the same conditions. The finer 
differences of reaction usually given in a final race analysis have therefore been 
ignored in this survey, and the terms immune, resistant, and susceptible adopted. 
Immunity and susceptibility are probably synonymous with complete incompatibility 
and complete compatibility respectively between host and pathogen. Resistance includes 
the intermediate range of reaction probably synonymous with intermediate incom- 
patibility. The use of the term seems quite valid, since such reactions less than 
completely susceptible generally agree with field resistance. 

Reaction Levels of the Differential Varieties. 
The following is a brief summary of the number of reaction levels determined for 
each of the differential varieties. 

Buda: 2: susceptible and a highly variable resistant reaction. 

WiLLisTON Golden: 2: susceptible and resistant. The resistant reaction tended 
towards immunity and was characterized by necrosis and a variable number of pustules 
rarely exceeding Flor's type "3-". A slight difference in reaction between races 1 and 17 
was too fine to justify differentiation of two intermediate reactions. 



BY H. B. KERR. 41 

Akmolinsk: 2: susceptible and resistant. The latter approximated fairly closely 
to Flor's type "1" reaction The susceptible reaction to race 1 may have been slightly 
lower in the scale of susceptibility than that evoked by other races. The variety 
appeared to have a measure of resistance to Punjab-attacking race 1 later in develop- 
ment, but in the seedling stage under optimal growing conditions it was fully 
susceptible. 

J.W.S.: 3: susceptible, resistant, immune. The resistant reaction to race 1 ranged 
from almost complete susceptibility to near immunity according to the conditions of 
incubation and the inoculation technique used. 

Abyssinian: 2: susceptible and immune. The susceptible reaction was sometimes 
depressed slightly to moderate susceptibility. 

Kenya: 3: susceptible, resistant, and immune. The resistant reaction was 
characterized by necrosis, type "2" and occasional type "3-" pustules. The susceptible 
/eaction was sometimes reduced to moderate susceptibility. 

Argentine 705-1: 2: a highly variable resistant reaction varying from near 
immunity to near susceptibility, and a rather variable susceptible reaction commonly 
depressed to moderate susceptibility. 

Very Pale Blue Crimped: 3: susceptible, resistant, and immune. The resistant 
reaction ranged from near immunity to type "2" reaction. The susceptible reaction 
was most commonly reduced to moderate susceptibility. It was very subject to 
variations in incubation conditions. 

Ottawa 770B: 2: susceptible and immune. Two very stable reactions. 

Argentine 462: 1: immune. 

Bison: 2: full susceptibility, immune. Both reactions were extremely stable. 

Punjab: 2: susceptible and immune. The susceptible reaction to the only race to 
*rnlch it succumbed was one of the most virulent noted during these studies. 

Walsh: 2: susceptible and resistant. The reaction to race 1 approximated more 
closely to immunity than that given by other non-Walsh-attacking races. The resistant 
reaction to these latter races was characterized by severe necrosis, with occasional 
type "2" pustules. The susceptible reaction was rather low in the scale of susceptibility. 

Morye: 1: immune. 

Newland: 2: susceptible and immune. Several races gave a delayed resistant 
reaction some time after completion of the usual incubation period. It was charac- 
terized by isolated and never abundant type "3" pustules. There was no necrosis. 
Uredospores taken from these pustules and built up on fully susceptible F257 failed 
to induce a more susceptible reaction. The pustules did not represent contaminant 
races. The races evoking this reaction more commonly gave complete immune reactions. 
The sparsity of type "3" pustules and the inconsistent development of the reaction did 
not justify differentiation of a resistant reaction level. 

Bolley Golden: 3: susceptible, resistant, and immune. The resistant reaction 
agreed with Flor's type "1", but was often depressed to immunity. Small isolated 
non-necrotic pustules like those recorded for Newland were sometimes associated with 
this reaction. 

Italia Roma: 2: susceptible and immune. There seemed to be a resistant reaction, 
but it was not sufficiently consistent to differentiate it from the slightly variable 
immune reaction. The susceptible reaction was quite commonly depressed to moderate 
susceptibility and occasionally to a resistant reaction. 

Leona: 2: susceptible, sometimes depressed to moderate susceptibility and a 
ilightly variable immune reaction. 

Tammes' Pale Blue: 2: susceptible and immune. Some immune reactions developed 
occasional type "3" pustules rather like those reported for Newland and Bolley Golden. 
The susceptible reaction varied about a norm of moderate susceptibility. 

Koto: 1: immune. 



42 



PHYSIOLOGICAL SPECIALIZATION OF MELAMPSOEA LINI IN AUSTRALIA, 





fc 






s 


q 


t-l 




n 




< 





n^ 


, 




CO d 
CO n. 


HHhHI— IWH-IHHI— II— (1— (WH-ll-HhHl— (HHHHC/J)— I 


CO l-H 


1— it-icCMCcajtBcocOf/jGOcCJ/ia^ajcci— ic/i' 


c^ pi 

CO l-H 




CO cq 


MhHC^C/2MHHKHWMhHI-I^C»HHhHpHl-HM 




Ml— IhHCOCOl— Ih-iM^I— (1— ()— iCOCOhHl— ihHhH 


00 -■ 






p5P^P^P^i:!5fi^P^2^P50^a^P5P^P^P5P5rie^ 










HHCccoc/icccoaiaDcccocococ/jcccococct/; 






o ^ 
^ o 


hHhHM)— iMi— icoi— iHHMi— iHHi— icocoajjyii— 1 


d 

o pq 


M(/)^c»cccoajPHa)^coPHaiw3c»cca2p(H 




Pnfi^pHpH^iKajpH^aiaja^cKccczj^icziai 


^^ 


MpHC»(/3«3P^C/:PHpHaiCOCCt»C»pH«;c»pH 


-i 


1— II— lOKl— liX'CCCCCOCOCOMiC/JCWCOaiCZJC/JC/:! 


K-5 


pHHHCO«5i72t»C/:a2«2ajCZlC/3^CCCBCZ;^C» 


-i 


C/:PCHC/jPHCCCOa)C/JCZ;iy2COC/J«3C/:i73COCOCO 


-i 


pHcocoa!cocccca!iKi^aDa:;^cca2«2PHCc 




PHC/jCCCZJajCCCZJCCCCCOCOCOCCCOCCC^f/}^) 


03 


i-<iMco-rtHioy:'t-xciOr-<(Mcor^ir;--Of--iX 



BY H. B. KEKK. 



43 



Races Identified During the Physiologic Race Survey. 

On the basis of the reaction levels determined for each variety eighteen races were 
detected among sixty-two accessions collected in the field. These races were numerically 
designated, instead of following the alphabetical system previously adopted by 
Waterhouse and Watson (1943). This was considered advisable for three reasons. The 
frequency of occurrence of new races would soon have exhausted the reservoir of 
English alphabet symbols. The differential series adopted in the current survey 
Included two varieties not used in any prior survey. The reaction given by Buda 
(used in previous surveys) to Punjab-attacking accessions was not considered suiHciently 
reliable to justify separation of races on this variety. The last two points complicated 
comparison of races differentiated in this and earlier surveys. A new system was 
therefore adopted to avoid confusion with results obtained in earlier surveys. 

The reaction of each race on the differential series is listed in Table 1. A bey 
for the identification of each race was also devised (Table 2). 



Table 2. 

Key jor the Identification of Australian Races oj Melampsora llni Determined in the Current 

Survey. 

Kace 



Punjab : Susceptible 
Punjab : Immune. 

Ottawa 770B : Immune. 
Walsh : Susceptible 
Walsh : Resistant. 
Newland : Immune. 
Very Pale Blue Crimped : Resistant. 
Argentine 705-1 : Resistant 
Argentine 705 -1 : Susceptible. 
Kenya : Resistant 
Kenya : Susceptible . . 
Very Pale Blue Crimped : Susceptible. 
Abyssinian: Immune 
Abyssinian : Susceptible. 
Bolley Golden : Immune 
BoUey Golden : Resistant. 
Argentine 705-1 : Resistant 
Argentine 705-1 : Susceptible 
Newland : Susceptible. 
Abyssinian : Immune 
Abyssinian : Susceptible. 
Bolley Golden : Immune. 
Kenya : Resistant 
Kenya : Susceptible . . 
Bolley Golden : Susceptible . . 
Ottawa 770B : Susceptible. 
Newland : Immune. 
Williston Golden : Resistant 
Williston Golden : Susceptible. 
Kenya : Resistant 
Kenya ; Susceptible. 
Italia Roma : Immune 
Italia Roma : Susceptible 
Newland : Susceptible 



11 



18 
12 



3 
10 



5 
13 



17 

15 

7 
16 

14 



Comparison of Races Identified During the Current Survey with Races 

Identified Earlier. 

Comparison of races identified in the present survey with those identified by 
previous workers was complicated by slight differences in the varieties used and by 
differences in the system of reaction classification. Pour different reactions were 
assigned to Kenya, Argentine 705-1, and Very Pale Blue Crimped by other workers. 
This was considered an unjustifiably fine distinction of reactions during this survey, 
and only three reactions were assigned to Kenya, and two to each of the last two 
varieties. Despite this, there was sufficient common ground for comparison. 



44 PHYSIOLOGICAL SPECIALIZATION OF MELAMPSORA LINI IN AUSTRALIA, 

Comparison with results obtained in the first survey by Waterhouse and Watson 
in 1941 and 1943 suggested a marked change in the rust population in Victoria and 
Western Australia. A more intensive survey would probably have revealed the same 
situation in South Australia. 

Eight of the twelve races identified from Western Australia since 1948 were 
distinctly different from races detected between 1940 and 1942. Among them five new 
races could be designated even on the restricted differential series used by Waterhouse 
and Watson. Races 2 and 4 were more avirulent on the original differential series 
used in Survey I than earlier non-Punjab-attacking races, and races 7, 14, 15, 16 and 17 
were highly virulent on Ottawa 770B, immune to all Australian races identified during 
the earlier survey. 

Six of the thirteen races identified in Victoria since 1948 were distinctly different 
from races identified in Survey I. Each of the six could be differentiated from each 
other on the original Survey I differential series. Races 2 and 4 were less virulent 
than non-Punjab-attacking races identified in Survey I. Races 8, 12 and 18 differed 
from Survey I races in their resistant reaction on Very Pale Blue Crimped, and 
race 14 was the first Ottawa 770B-attacking race recorded in the State. 

Fewer collections were received from South Australia, and correspondingly fewer 
races were identified. But one of the five identified since 1948, race 2, differed from 
the four races recorded in 1943. 

The race position remained unchanged in New South Wales, where both 
collections received were Punjab-attacking. 

No collections were received from Queensland prior to 1948. The seven received 
and analysed since then were identified as race 1, a Punjab-attacking race. 

A marked shift in the race population was noted when races identified by 
Watson, Charles, and Kerr between 1943 and 1948 were compared with races identified 
in the current survey. Prior to the current survey, only three Newland-attacking 
races had been identified. There were no Bolley Golden or Ottawa 770B-attacking 
races, and Walsh was relatively free from infection in the field until 1948. This 
survey identified six Newland-attacking races, 4, 5, 9, 12, 13 and 14, each of them 
apparently different from already determined Newland-attacking races. Five Ottawa 
770B-attacking races, 7, 14, 15, 16 and 17, and three Bolley Golden-attacking races, 
4, 12 and 13, were detected. An important Walsh-attacking race, race 11, was 
discovered in Victoria. 

Only four of the races identified in this survey bore any close resemblance to 
races identified by previous surveys. Races 6 and 10 agreed fairly closely with 
Charles' races C and E respectively. Race 2 was identical with race K. The latter 
race was identified by Kerr in 1948. It was later found to give slightly different 
reactions on J.W.S., Akmolinsk, Kenya, and Argentine 705-1 than those listed by 
MilMkan (1951). Race 1 was probably a composite of races A and F, since the reaction 
of Buda was not used to separate races in this survey as it had been to differentiate 
races A and F in earlier surveys. 

Races A and F were differentiated by their reaction on Buda, and since this 
variety was rejected during the current survey, differentiation between races A and F 
was impossible. Race 1 cannot be equated with either of the former races and 
must be equated with both. Its reaction differs somewhat from that indicated for 
races A and F on two of the differential hosts. This could be attributed to the 
inoculation technique adopted in this survey. This induced a more virulent reaction 
than the spore dusting method of earlier surveys. When race 1 was inoculated onto 
J.W.S. and Akmolinsk by the latter method it gave virtually the same reactions, 
immune and resistant respectively, as those noted by Waterhouse and Watson for 
races A and F for these two varieties. 

It was concluded that fourteen of the eighteen races identified in this survey had 
not been recorded before in Australia, namely races 3, 4, 5, 7, 8, 9, 11, 12, 13, 14. 15, 
16, 17, and 18. Some of these races may have been present in the field when earlier 



BY H. B. KERR. 



45 



surveys were carried out, but may have been overlooked because insufficient districts 
were sampled, or because they constituted only a very small part of the total rust 
population at that time. But the very high percentage of new races among those 
identified in the current survey suggests a marked change in the race complex 
since the first surveys were carried out. 



Table 3. 
Distribution of Australian Races in Time and State. 



State. 


Year. 
















Race. 




















Total. 


1 


2 


3 


4 


5 


6 


7 


8 


9 


10 


11 


12 


13 


14 


15 


16 


17 


18 


Western 
Australia. 


1918 
1949 
1950 
1951 


1 


1 


— 


1 


2 
3 


— 


1 


- 


1 


— 


— 


— 


1 
2 


2 


1 


1 


1 


1 


4 
2 
5 
8 


South 
Austraha. 


1948 
1950 
1951 
1952 


1 
1 

2 


1 
1 


1 


— 


1 


- 


— 


— 


1 


— 


— 


— 


— 


— 


— 


- 


— 


3 
2 
2 
2 


Victoria. 


1948 
1949 
1950 
1951 
1952 


1 


1 
2 


1 
2 


1 
1 


2 
1 


1 
1 


- 


2 


1 


1 
1 
2 


1 


1 


- 


1 


— 


— 


— 


1 


4 
11 
3 
2 
5 


Queens- 
land. 


1948 
1952 


1 
6 


— 


— 


~ 


— 


— 


— 


— 


- 


— 


— 


— 


— 


- 


— 


— 


- 


— 


1 
6 


New 
South 
Wale^. 


1949 
1952 


1 

1 


— 


— 


— 


— 


— 


- 


- 


- ■ 


— 


— 


— 


- 


— 


— 


— 


— 


— 


1 
1 


Total of 
each race 




15 


G 


4 


3 


9 


2 


1 


2 


2 


5 


1 


1 


3 


3 


1 


1 


1 


2 


02 



Races Detected in Each State Durhis tlie Survej . 



Queensland 

New South Wales . 

Victoria 

South Austraha 

Western Australia . 



Race 1. 

Race 1. 

Races 1, 2, 3, 4, 5, 6, 8, 9, 10, 11, 12, 14, 18 

Races 1, 2, 3, 5, 10. 

Races 1, 2, 4, 5, 7, 9, 13, 14, 15, 16, 17, 18. 



Geographic Distribution of Each Race and the Major Race Groups. 

The distribution of the eighteen races according to the year of collection and State 
is given in Table 3. The districts in each State in which each race was found and a 
summary of the race complex of each State are listed in Table 4. 

There was a remarkable diversity of non-Punjab-attacking races in the flax-growing 
southern States, and a complete absence of any of these races in the eastern linseed- 
growing States, New South Wales and Queensland. 

There was no one dominant race among the seventeen non-Punjab-attacking races 
found in the south, although races 2, 5 and 10 occurred with a slightly greater 
frequency than the others. This follows the pattern of Flor's findings in the United 
States. Only two non-Punjab-attacking races, 2 and 5, occurred in all three southern 
States. Nine races were restricted to a single State. 

Newland-attacking races were widely distributed in the southern belt during this 
s-urvey, but were already established before the survey. Ottawa 770B-attacking races 



/46- PHYSIOLOGICAL SPECIALIZATION OF MELAMPSORA LINI IN AUSTRALIA, 

by contrast were recorded in Australia for the first time since this survey began. 
There was no trace of such a race among the four hundred or more collections 
analysed previously. In 1948 Western Australian departmental officers reported sparse 
infection of Ottawa 770B seedlings in a small experimental plot at Kojunup. They 
attributed it to varietal impurity. The first Ottawa-attacking race, race 7, was 
identified in 1949 in a collection from Boyup Brook. Another Ottawa-attacking race, 
race 14, was identified from the same district in 1950. Three other Ottawa-attacking 
races, races 15, 16 and 17, were discovered in 1951 in the same area at North Boyup 
Brook, Newbicup, and Kulikup respectively. No further collections were received from 
the State, but a member of the local Department of Agriculture wrote that "The 

Table 4. 
Details of the State and District of Occurrence of the Races Determined durimj this Survey. 
Race 1. Queensland ; Gatton 1948, Hermitage and Toowoomba 1952, Biloela 1950.* 
N.S.W. : Castle Hill 1949, Tibooburra 1952, Curlewis 1950.* 
Victoria : Lake Bolac 1949, Werribce 1953.* 
South Australia ; Waite Institute 1950, 1952, Corry Point 1951. 
Western Australia : Boyup Brook 1950. 
Race 2. Victoria : Lake Bolac 1948, Winchelsea, Colac 1949. 
South Australia : Laura 1948, Waite Institute 1951. 
Western Australia : Boyup Brook 1949. 
Race 3. Victoria : Winchelsea 1949, Thorpdale 1952. 

South Australia : Maclaren Vale 1948. 
Race 4. Victoria : Werribee 1949, 1950. 

Western Australia ; Mayanup 1948. 
Race 5. Victoria : Myrtlclord 1948, Lake Bolac 1948, 1949. 
South Australia : Undalaya, 1948. 

Western Australia: Kojunup 1948, North Boyup Brook, Beiijiiiup 1951. 
Victoria : Colac 1948, Winchelsea 1949. 
Western Australia : Boyup Brook 1949. 
Victoria : Lismore, Casterton 1949. 
Victoria : Winchelsea 1949. 
Western Australia : Boyup Brook 1950. 
Victoria : Colac 1950, Drouin 1951, Thorpdale 1952. 
South Australia : Laura 1950. 
Victoria : Casterton 1949. 
Victoria : Birregurra 1950. 

Western Australia: Boyup Brook 1950, Mayanup 1951. 
Victoria : Thorpdale 1952. 
Western Australia : Boyup Brook 1950. 
Western Australia : North Boyup Brook 1951. 
Western Australia : Newbicup 1951. 
Western Australia: Kulikup 1951. 
Western Australia : Kojunup 1948. 
Victoria : Lismore 1951. 

* Not positively identified as race 1 at Sydney University, since no viable inoculum 
was received, but definitely Punjab-attacking races, and therefore closely related to race 1, 
if not actually race 1. 

Ottawa-attacking race (group) was very active (in 1952), and will, I think, considerably 
lower the fibre value of commercial crops of Wada". It seemed to have supplanted 
the old races. Until then the race group had been confined to Western Australia, 
but in early December. 1952, race 14 was identified on a specimen of Rust Resistant 
Norfolk Earl growing in Government experimental plots at Thorpdale, Victoria. 

Field infection of Walsh, the only commercially grown linseed variety in Australia, 
was relatively insignificant until 1948. Although Walsh had not yet been included in 
the differential series, it is fairly certain from the above observation that no major 
Walsh-attacking race was then present. Race 11 was discovered in a collection from 
Casterton, Victoria, in 1949. It was highly virulent on the rather mixed commercial 
variety, attacking 70% of the seedlings in glasshouse tests at Sydney. The single 
plant selection, previously resistant to already identified Australian races, was fully 
susceptible. Despite its virulence and an abundance of susceptible host material in 
the field, it did not appear to have spread. But since the conclusion of this survey 



Race 


6. 


Race 


7. 


Race 


8. 


Race 


9. 


Race 


10. 


Race 


11. 


Race 


12. 


Race 


13. 


Race 


14. 


Race 


15. 


Race 


16. 


Race 


17. 


Race 


18. 



BY H. B. KERR. 47 

there has been a report of a new Walsh-attacking race at Thorpdale and Casterton 
(Debrett, 1954). 

Bolley Golden-attacking races were found in Victoria and Western Australia. 
Race 17, already mentioned as an Ottawa-attacking race, was the first occurrence of a 
Tammes' Pale Blue-attacking race in Australia. It was located at Kulikup in Western 
Australia. 

Race 1 was the only Punjab-attacking race identified during the survey. It was 
the most common of the eighteen races and v/as detected in fifteen of the sixty-two 
collections. It occurred in all five States from which the collections were received. 
It was most common in Queensland. It was less common in New South Wales, Victoria, 
South Australia and Western Australia. But there was ample evidence from information 
supplied by officers of the State departments of agriculture and in the literature to 
indicate a more widespread occurrence. 

Two Punjab-attacking accessions from New South Wales were identified as race 1. 
One was obtained from Tibooburra on wild flax, L. m.arginale. The other was obtained 
from linseed experimental plots at Castle Hill, near Sydney. Both districts were 
hundreds of miles from any known source of infection from commercial varieties of 
flax or linseed. Race 1 or a closely related race was also present in experimental 
plots at Curlewis in the north-western wheat belt. Several flax and linseed varieties 
were sown at the University Experiment Station, Curlewis, in May, 1950. The seed 
vas disease-free, and there was no known source of rust inoculum in any of the 
commercial crops grown within 500 miles of the district. Punjab and Imperial, which 
in tests at Sydney University were immune to every race identified in this survey 
except race 1, were heavily infected by the end of September. The spores received on 
specimens sent from the district were inviable on receipt, but the race was almost 
certainly race 1 or a closely related race. The other varieties, including Concurrent, 
were uninfected. Since this variety was fully susceptible to all the non-Punjab-attacking 
races identified during this and previous surveys the immunity of this variety at 
Curlewis was strong proof of the absence of non-Punjab-attacking races from the 
district. 

Rust was also noted on L. marginale at Pucuwan, 11 miles from Temora in the 
south-western wheat belt, during a field trip in November, 1948. Mr. Evans, of the 
Sydney University Botany Department, also reported that rust was quite common 
on the species near Sydney. Since twelve of thirteen collections received from the 
species since 1941 were Punjab-attacking races it is reasonable to infer that Punjab- 
attacking races were present near Pucuwan in 1948 and are ccmmon near Sydney in 
most years. 

Only one collection of race 1 was received from Victoria during this survey, but 
seed of the differential varieties and F257 was sent to Mr. P. Debrett, of the Victorian 
Department of Agriculture, in 1952. Owing to excessive rain he was unable to make 
sowings in the Werribee district until late in the year. Punjab became heavily infected 
iuring February. The other varieties were immune. Since no inoculum was forwarded, 
the exact nature of the race is unknown. But since it was a Punjab-attacking race, 
it was probably race 1 or a closely related race. Since Bison and F257 were not 
infected, none of the common non-Punjab-attacking races could have been present in 
the field at that time of the year. 

Four collections of race 1 were received from South Australia, one from L. marginale 
growing at Corry Point in 1951, and three from Punjab-type Indian linseed varieties 
growing at Waite Agricultural Institute. The race was present each year from 1950 
to 1952 and constituted four of the ten collections received from the State. 

One collection of race 1 was received from Western Australia. It was found as a 
mixture with race 14 on supposedly Concurrent plants growing at Boyup Brook in 
1950. It must have been more common, since fifteen of twenty-eight rust collections 
sent in from the State in 1942 were identified as race A by Waterhouse and Watson. 

The first Punjab-attacking collection from Queensland was collected from linseed 
growing at Gatton in 1948. No further collections were received from the State until 



4,S PHYSIOLOGICAL SPECIALIZATION OF MELAMPSORA LINI IN AUSTRALIA, 

1952, but during that year special efforts were made to determine whether race 1 was 
the only race present in Queensland. 

Records were kindly forwarded by the Pla7it Introduction Officer of the Common- 
wealth Scientific and Industrial Research Organization at the Cooper Laboratory, 
Lawes, Queensland, giving the mean intensity of rust infection in this district of the 
differential varieties used by Waterhouse and Watson and of Liral Crown and 
Concurrent. Results obtained in 1948 and 1949 showed that only Punjab-attacking 
races were present. Tests were discontinued in later years, but up to the end of 
this survey there was no record of infection of varieties such as Bison and Concurrent, 
known to be susceptible to all of the non-Punjab-attacking races of the southern States. 

Seed of the varieties Punjab (susceptible only to race 1) and F257 (susceptible to 
all the races, including race I) was distributed to experimental stations at Warwick, 
Hermitage, Toowoomba, and Kingaroy in 1952. They were grown under conditions 
most likely to induce infection, but remained uninfected at Kingaroy and Warwick. 
Infection of both varieties was reported at Hermitage in mid-November. The rac( 

Table .5. 
Distribution of Major Race Groups in Australia. 
Punjab-attacking Race Group: Bace 1, 

Queensland : Gatton, Hermitage, Toowoomba, Biloela. 

N.S.W. : Castle Hill, Tiboobuira, Curlewis. 

Victoria : Lake Bolac, Werribee. 

South Australia : Waite Institute (Adelaide), Corry Point. 

Western Australia : Boyup Brook. 
Ottawa rroB-attackiny Races : 7, 14, 15, 16 and 17. 

Victoria : Thorpdale. 

Western Australia : Boyup Brook, North Boyup Brook, Kulikup, Newbicup. 
Newland-attacking Races: 4, 5, 9, 12, 13 and 14. 

Victoria : Winchelsea, Lake Bolac, Birregurra, Werribee, Myrtleford, Thorpdale. 

South Australia : Undalaya. 

Western Australia : Mayanup, Kojunup, North Boyup Brook, Benjinup, Boyup Brook. 
Bolley Golden-attacking Races: 4, 12 and 13. 

Victoria : Birregurra, Werribee. 

Western Australia : Mayanup, Boyup Brook. 
Walsh-attacking Race : 11. 

Victoria : Casterton. 

was not identified, but three collections received from the same district between 
September and October were identified as race 1. 

The varieties were sown at three localities in the Toowoomba district. Sowingi 
were made at regular intervals and a close check kept on each plot. No rust was 
recorded at one of the locations. Both varieties were infected in early December at 
the other locations. F257 was moderately infected. Punjab was heavily infected. 
The rust was identified as race 1. Sowings were made on 6th January, 1953, and at 
fortnightly intervals for some time afterwards. Every attempt was made to induce 
healthy succulent growth to encourage infection, but no rust was recorded on either 
variety after December. 

Rust has appeared at Biloela Experiment Station, 560 miles north of Brisbane. 
This was almost certainly race 1 or a related race, since rust specimens sent to 
Sydney from the area in October, 1950, were from Punjab and Imperial. 

Notes on the Cultivated and Wild Hosts in Australia. 

Concurrent was highly susceptible in the seedling stage to all the races identified 
during the survey. It was somewhat less susceptible to race 1 than the other races 
in the seedling stage, and infection was usually restricted to the leaves, rarely 
spreading to the stems. Later in maturity it seemed to acquire complete immunity 
to the Punjab-attacking races. The same mature plant resistance seemed to be 
effective against the new highly virulent Ottawa-attacking races of Western Australia. 
This was not apparent until 1951. The first Ottawa-attacking race, race 7, was 



BY H. B. KERB. 



49 



collected at Boyup Brook from the variety Boyup In 1949. The following year race 14 
was identified from two collections of Concurrent. Eight collections were received 
from neighbouring areas in 1951. The three collections from Concurrent yielded 
non-Ottawa-attacking races. Three different Ottawa-attacking races were isolated from 
the Wada specimens. This suggested that Concurrent was resistant to the most recent 
Ottawa-attacking races. The Western Australian Department of Agriculture late in 
1952 reported that Concurrent was then free of rust although Wada was seriously 
affected. 

Excised shoots of Concurrent were inoculated and incubated with an Ottawa- 
attacking race, race 17, and a non-Ottawa-attacking race, race 6. The two sets of 
shoots were kept under identical conditions in the same part of the glasshouse until 
sporulation and for several weeks afterwards. Both sets of shoots became heavily 
infected within two weeks. Shortly afterwards the race 6 infection spread to the 
stems, and subsequent shoot growth was stunted. Race 17 infection was restricted to 
the leaves, and the shoots recovered and continued vigorous growth (Plate iv, 2). 



Table 6. 
Races Isolated from the Varieties and Specii 



Listed Below. 



Host. 






Races. 


Linum usitatissimum : 








Concurrent . . . . 




1, 


4, 5, 13, 14. 


Liral Crown 




2, 


3, 5, 8, 10. 


Walsh 




3, 


6, 8, 11, 12, 18. 


Wada 




9, 


10, 13, 15, 16, 17. 


Boyup (A.3) 




3, 


5, 7, 13. 


Punjab, Imperial 




1. 




Unidentified flax varieties, probably Concurrent or 


Liral 






Crown 




2 


4, 5, 10, 18. 


Linum marginale 




1. 





Wada and Boyup were developed as rust-resistant varieties to replace Concurrent 
and Liral Crown. Tests at Sydney University showed that both varieties carried an 
appreciable percentage of off-type plants susceptible to most of the non-Punjab-attacking 
races with which they were tested. Both were fully susceptible in the seedling stage 
to the Ottawa-attacking races. But Boyup, like Concurrent, from which it is reported 
to have derived by natural crossing with Wada, seemed to have mature plant resistance 
to these races in the field. The two collections received from Boyup, Western 
Australia, in 1951 yielded non-Ottawa-attacking races, and a letter from the Department 
of Agriculture in November of 1952 stated that it was not as badly infected as Wada 
in the field. 

The commercial variety Walsh contained up to 20% seedlings susceptible to many 
of the non-Punjab-attacking races. The occurrence of these susceptible plants probably 
gave rise to the many reports of important new Walsh-attacking races from the 
southern fiax-growing districts. Race 11 was the only race to which most Walsh 
seedlings were fully susceptible, but eight of twenty-four single plant selections from 
the commercial variety were immune to this race. The occurrence of race 11 at 
Casterton in 1949 posed a considerable threat to the linseed-growing industry, but 
up to 1952 no further specimens of this race were received from Victoria. A report 
from Victoria in 1954 would seem to indicate the appearance of a new Walsh-attacking 
race in the Casterton and Thorpdaie districts. 

Linum viarginale is probably an important host of M. Uni in Australia. It is 
native to Australia and has been a host to M. Uni since the first recorded observations 
of rust in the country. (The species is common in New Zealand, where it is regarded 
as an important carry-over host, particularly during winter. New Zealand crops are 
spring sown. Unlike Australia, there is an abundance of cultivated host material 
during summer, but a dearth of the host during winter.) It was reported from 



50 



PHYSIOLOGICAL SPECIALIZATION OF MELAMPSOEA LINI IN AUSTRALIA, 



Western Australia: "The evidence is clear in this State that L. marginale is an 
Important hold over host for the rust fungus." The species is quite common in 
Victoria and South Australia. It has been found at such widely separated places in 
New South Wales as Sydney, Bourke, Curlewis, Tibooburra and Armidale. 

It appeared to be much more restricted in its occurrence in Queensland. One 
report indicated that it had never been seen in the Lawes district or on the Darling 
Downs. The Government Botanist reported that the species was not common and 
the four localities from which specimens had been received were in the southern 
portions of the State towards the New South Wales border. 

The species is not homogeneous in its reaction to M. lini. During the course of 
these studies, seed of the species was received from several States. The different 
lines were sown and tested for reaction to race 1 and several of the non-Punjab- 
attacking races. The lines tested were immune to several accessions of race 1 and 
also races 2, 6 and 7. Other lines, however, have been susceptible to race 1. 

Factors Affecting the Incidence of the Rust in the Field. 

The yearly and seasonal incidence of rust was directly correlated with the 
availability of susceptible material. Flax acreages have diminished steadily since the 
last war from 61,000 acres in 1944 to approximately 12,000 acres in 1948, and approxi- 
mately 4,800 in 1950. This was paralleled by a reduction in the number of collections 
of rust received from the field. 

Most of the crops were harvested by the end of December, leaving a sparse supply 
of volunteer plants to carry the rust over the hot summer months of January and 
February in the uredospore stage. This partly accounts for the fact that few samples 
of rust were received from the field later than December. Details of the earliest and 
latest collections of rust received from the field are given in Table 7. The dates given 

Table 7. 

Details of the Earliest and Latest Collections of Rust Received from Each State. 
1948-1952. 



State. 


Earliest 


Latest 




Collection. 


Collection. 


Queensland 


8th September. 


13th December. 


New South Wales 


26th September. 


22nd November. 


Victoria 


3rd August. 


23rd December. 


South Australia 


9th October. 


15th January. 


Western Australia 


10th September. 


4th December. 



in this table do not indicate the earliest or latest appearances of M. lini in the field 
in the uredospore stage, but do help to indicate roughly the periods of maximum 
development of the rust in the uredospore stage. 

The rust is evidently most abundant in the field in the uredospore stage from 
early September to December. Lack of rust before this period can be attributed to low 
field temperatures. 

Observations in the field and in the glasshouse at Sydney and reports from 
workers in other States indicate that Punjab-attacking race 1 is better able to survive 
higher summer temperatures than the non-Punjab-attacking races in the uredospore 
stage. Non-Punjab-attacking race 2 was unable to survive summer temperatures in 
the field at Sydney. But Punjab-attacking race 1 survived without diSiculty and caused 
severe infection. 

The high temperature tolerance of Punjab-attacking races was also apparent from 
Mr. P. Debrett's report on the infection of varieties forwarded to him at Werribee in 
1952. Punjab was severely infected during February, 1953. Non-Punjab-attacking races 
were much more common than Punjab-attacking races in his district up to December 



BY H. B. KERR. 



51 



in earlier years. But they were unable to maintain themselves in the uredospore stage 
during late summer, since Bison and F257, highly susceptible to these races, were 
uninfected during February. 





80°- 


■K 




75 


Pertt^ 




e 


nd^^o 


Text-fig. 2.- 


— Clir 


growing 


states. 






Climatic map of Australia indicating mid-summer isotherms of flax and linseed 



Approximately 1,000 miles further north, at Toowoomba, even Punjab-attacking: 
races were less able to survive summer in the uredospore stage. Race 1 was present 
in the field until late December but failed to cause infection of susceptible material 
during January and February. 

Table 8. 
Summer and Winter Temperatures of Potential Flax Growing Districts in Australia. 







Mean 








Temperature 


Mean Temperature 


State. 


District. 


during 


during Summer. 






Mid Winter. 
°F. 


"F. 


Tasmania. 


Launceston 


47 


64 


Western Australia 


Bunbury 


55 


73 Dec-Jan. 




Katanning 


50 


70 Dec-Jan. 




York 


50 


76 Dec-Jan. 


South Australia. 


Waite Institute 


50 


70 Dec-Feb. 




Millicent, Kapunda 


Much the sam 


3 as Waite Institute. 




Mt. Barker, Clare 






Victoria. 


Colac 


43 


64 Dec-Feb. 




Leongatha 


44 


64 Dec-Feb. 




Wangaratta 


44 


72 Dec-Feb. 




Myrtleford 


38 


70 Dec-Jan. 


New South Wales 


Albury 


44 


73 Dec-Feb, 




Cootamundra 


38 


72 Dec-Feb. 




Glen Innes 


38 


68 Dec-Jan. 




Inverell . . 


44 


73 Dec-Jan. 




Bathurst 


38 


68 Dec-Jan. 



Although non-Punjab-attacking races lacked tolerance to high temperature they 
were able to survive in the field in the uredospore stage in some districts in the 
southern States when the host was available. Rusted Walsh plants were received 



52 



PHYSIOLOGICAL SPECIALIZATION OF MELAMPSORA LINI IN AUSTRALIA, 



from Mount Gambler in South Australia in m.id-January, 1953. The rust was inviable 
on receipt. But since Walsh is completely immune to race 1, the only Punjab-attacking 
race identified during this survey, it is safe to attribute the infection in this instance 
to a non-Punjab-attacking race. 

Since summer temperatures played an important role in regulating the seasonal 
incidence of rust, efforts were made to determine summer temperatures of the flax- and 
linseed-growing areas of Australia. 



Tabid 9. 
Summer Temperatures in Western Australia Flax Growinj Districts. 







Average of 40 years. 


°F. 




Montli. 


Bridgetown. 


Katanning. 
Mean. 


York. 
Mean. 




Mean. 


Maximum. 


Hottest Day. 


November . . 

December 

January 

February 

March 


61-9 
66-4 
69 1 
68-6 
65-3 


76-4 
82-3 
85-5 
85-3 
80-4 


103 
105 
109 
115 
105 


63-7 
68-5 
71 1 
70-6 
66-8 


67-8 
73-7 
76-4 
76-1 
71-7 



Climatic graphs prepared by Dr. Forster (1941) to determine suitable areas for 
flax production in Australia yielded the information found in Table 8. Additional 
information was furnished by the Western Australian Department of Agriculture for 
Western Australian flax areas (Table 9). Temperatures at representative areas on the 
Darling Downs and other districts in Queensland were obtained from S. G. Gray's 
report on variety trials, 1950 (Table 10). 

These figures agreed with the climate maps indicating the major temperature zones 
in Australia during the summer. It will be noticed that none of the Queensland 
districts had a mean temperature during January of less than 71 °F. Some at least 
of the flax-growing districts of the southern States had a mean temperature during 
January of 70°F. or less. This fact is highlighted by the January (mid-summer) 
isotherms. Most of the southern flax-growing areas of Victoria, South Australia and 
Western Australia lie in a moderately temperate zone by comparison with the linseed- 
growing areas of north-west New South Wales and Queensland. 

Table 10. 
Mid Summer and Mid Winter Temperatures in Queensland Linseed Growinn Districts. °F. 



Month. 


Mean Monthly Temperature. 


Lawes. 


Toowoomba. 


Pittsworth. 


Killarney. 


Nanango. 


January 

July 


77-2 
561 


71-2 
50-7 


74 1 
51-4 


71-9 
59-0 


74-1 
51-2 



Additional Comments on the Virulence of Punjah-Attacking and IS! on-Punjah- Attaching 

Races. 
The results already summarized indicate an important division within the M. lini 
complex between the Pur.jab-attacking and the non-Puujab-attacking races. Observa- 
tions not directly relevant to the survey were made during the course of these studies, 
which conflrmed the difference between the pathogenicity of the two race groups. 

It was possible within the reaction level of susceptibility to distinguish varying 
levels of virulence. Such distinctions could be readily defined by noting the following. 



BY H. B. KERR. 



53 



1. The extent of infection. In cases of milder susceptibility the uredosori were 
entirely or almost entirely restricted to the leaves, seldom spreading to the stems. In 
cases of extreme susceptibility the uredosori developed as vigorously on the stems as 
on the leaves. 

2. The extent of plant recovery. In cases of mild susceptibility the inoculated 
plants commonly recovered and resumed normal growth unless deliberate attempts 
were made to induce further infection. In cases of extreme susceptibility plant growth 
was severely retarded or completely checked. Natural secondary infection of new 
growth was common. 

3. Varieties showing the most severe susceptible reactions commonly developed 
heavy infection by the spore dusting technique. Varieties giving a milder reaction 
developed a very sparse infection by the spore dusting method. 

"Without exception non-Punjab-attacking races evoked a maximally virulent sus- 
ceptible reaction on the variety F257. This variety was also susceptible to race 1, 
but invariably gave a less virulent susceptible reaction. A quantitative assessment of 
the difference in virulence of the non-Punjab-attacking and the Punjab-attacking races 
on F257 is summarized in Table 11. With the majority of the seedlings inoculated 



Table 11. 
Stem Infection of Pot Grown Seedlings Inoculated 7th May, 19.53. Read 29th May, 1953. 







Number 


of Stems witli Uredosori. 


Bace. 


Variety. 




Heavy 


Very Mild 


No 






Infection. 


Infection. 


Infection. 


1 


F257 


2 


18 




1 


Punjab 


19 


— 


4 


6 


F257 


12 


— 


5 


2 


F257 


10 


— 


5 



with NPA races infection spread to the stem, producing prominent uredosori from 
which considerable quantities of inoculum could be drawn. Only two of the twenty 
seedlings inoculated with race 1 developed many stem uredosori. Eight stems were 
scarcely infected, and ten stems developed only a few scattered pustules. Race 1, 
however, gave a reaction of maximum virulence on Punjab seedlings, nineteen of 
twenty-three seedlings developing a heavy crop of uredosori on the stem. 

In another test with excised shoots the milder virulence of race 1 on F257 was 
reflected in the almost complete lack of natural secondary infection after sporulation 
of the primary infection. Only one shoot shov/ed any appreciable secondary infection. 
Only six secondary pustules developed on the other twenty-five shoots. The extreme 
virulence of the same race on Punjab resulted in heavy secondary infection of twenty- 
one of the thirty-four shoots. 

Spore dusting bulk inoculation resulted in heavy infection of Punjab plants dusted 
with uredospores of race 1, placed in a moisture chamber and sprayed at intervals. 
F257 plants treated simultaneously in the same chamber developed only a light 
infection (Plate iv, 1 and 3). 

The susceptibility of Punjab to race 1 was one of the most virulent reactions noted 
during the course of these studies. Punjab-attacking races from New Zealand induced 
a much milder reaction. Australian race 1 and New Zealand Punjab-attacking races 3 
and 13 were cultured on Punjab excised shoots. The Australian race 1 accessions 507, 
610, 613 and 624 from different sources in Queensland, and accession 621 from South 
Australia induced a maximum susceptible reaction with typical type 4 pustules. 
Infection spread to the stems; the plants became stunted and did not resume normal 
growth. Shoots inoculated with the New Zealand races developed type 3 pustules. 



54 



PHYSIOLOGICAL SPECIALIZATION OF MELAMPSORA LINI IN AUSTRALIA, 



After three to four weeks the infected leaves died and normal growth resumed. 
Infection did not spread to the stems (Table 12). 

Table 12. 
Comparison of Virulence of Australian and New Zealand Punjab-attacking Races. 





Race 1 Accessions. 


New Zealand Races. 




507 


610 


613 


621 


624 


3 


13 


Stems with 
uredosori . . 


51/53 


9/9 


8/8 


11/11 


7/9 


0/18 


0/19 



A Note on Uredospore Morphology. 

Measurements were made of uredospores of races 1, 2, 6 and 7 over a period of 
two weeks. The spores were collected from susceptible plants cultured under identical 
conditions and stored at 0° to 2°C. in the refrigerator for no more than four weeks. 
The spores were prepared for measurement by dusting onto a drop of specially prepared 
lacto-phenol on a thoroughly clean microscope slide. They were covered lightly with 
coverslips and checked after at least half an hour under the oil immersion objective. 
Provided all excess lacto-phenol was blotted up and the immersion oil was very fluid 
the coverslip remained firmly fixed to the slide as the slide was moved under the 
objective. The tube of the microscope was extended until each division on the scale 
in the eyepiece corresponded to 1/j.. At this level of magnification the image of the 
spores was clearly resolved. The spore concentration was adjusted beforehand so that 
the spores were fairly widely separated from each other. The movement of the slide 
was manipulated by a mechanical stage. To avoid biased selection of spores all the 
spores were measured which appeared completely within the field in the movement 
of the slide from left to right. Several samples of uredospores of each race were 
studied and one hundred spores were measured per sample. 

Differences between the spores of different races were never significantly greater 
than differences between samples of the same race. Results obtained with one sample 
of spores are given in Table 13. 

Table 13. 
Uredospore Morphology (in ix) of Three Races Cultured under Identical Conditions and Collected 

23rd May, 1953. 
(Measured 23rd May, 1953.) 



Race. 


Average 
Length. 


Range. 


Average 
Breadth. 


Range. 


1 
2 

■ 7 


21-4 
20-7 
20-6 


(18-25) 
(16-24) 
(16-25) 


18-5 
18-0 

17-8 


(15-21) 
(1.5-20) 
(15-20) 



Discussion. 

The most striking feature of the race survey was the dichotomy of the Australian 
Melampsora lini race complex into the Punjab-attacking (PA) and non-Punjab-attacking 
(NPA) groups. This division paralleled differences in temperature tolerance in the 
field, and in vitro in the uredospore stage (Kerr, 1958), differences in host range in 
the field and in glasshouse tests, differences in the continuity of individual race 
members within each group in time and space. 

None of the NPA races had so restricted a host range as PA race 1. Race 2, the 
least virulent of the NPA races, was highly virulent on four of the twenty differential 



BY H. B. KERR. 55 

varieties. All the other NPA races were highly virulent on at least seven of the 
differential varieties. PA race 1 attacked only two of these varieties. J.W.S. gave an 
immune or fully susceptible reaction to the NPA races. Race 1 gave a variable 
resistant reaction even more marked in hybrids involving this variety. Bison was 
completely susceptible to all NPA races, as well as all races in America and New 
Zealand. It was immune to race 1. 

The division between the PA and NPA groups was not an arbitrary classiflcatory 
division such as existed between the Ottawa-attacking and non-Ottawa-attacking races 
of Western Australia, determined by a single pathogenic factor. It seemed to be an 
important natural division. This might suggest that the difference was physiologic 
and that incompatibility prevented intergroup hybridization. This was ruled out by 
Charles's report (personal communication) of hybridization between races of the two 
groups in the glasshouse at Sydney University in 1947. There was no appreciable 
difference in uredospore morphology to indicate subspecific difference. 

Hybridization between the two groups with progressive breakdown of the present 
differences might therefore have been expected. But there was no evidence that this 
had occurred since the first survey in 1940. Hybridization should eventually have 
produced among a wide range of races, races virulent on both Bison and Punjab, and 
others avirulent on both varieties. Since single factors condition virulence on these 
two varieties, the chances against the appearance of such races following hybridization 
were by no means great. Other widely virulent races, avirulent on Bison, and 
Punjab-attacking races lacking the high temperature tolerance of PA race 1 should 
also have appeared. The absence of such races from the field between 1940 and 1952 
confirms the view that the PA and NPA race groups exist as two non-interpenetrating 
closed systems. 

The existence of the two-group system poses several questions. How did the two 
systems originate? How is the dichotomy maintained and how long is it likely to 
continue? The answer to the second query is likely to suggest a solution to the first. 
If the difference did not stem from physiologic incompatibility it must have been 
maintained by factors external to the pathogen. Host and non-host environmental 
factors must have maintained the balance between the two groups. 

The two groups did not occur together in the field on the same host varieties. 
The NPA group parasitized the commercial fiax crops and were isolated on a number 
of occasions from the commercial linseed variety Walsh. The PA group, with one 
exception, was isolated from Linum, marginale, a wild host, or such linseed varieties 
as Punjab and Imperial, which were not grown commercially and which were immune 
to all the NPA races. One collection of PA race 1 was received from Concurrent, a 
common host to the NPA group, but the plant yielding the race may well have been 
a rogue. This variety, fully susceptible to most NPA races from the seedling to the 
mature plant stage, was susceptible to PA race 1 in the seedling stage only. The 
difference in host range may have militated against hybridization of the two race groups. 

The geographic distribution of the NPA group would seem to have been regulated 
primarily by temperature and to a lesser extent by host. There was an abundance of 
susceptible host in the southern flax-growing States. In the north the only commercial 
host was the linseed variety Walsh. Since this variety was generally resistant to the 
NPA races this might have militated against the epidemic development of the common 
NPA races in New South Wales and Queensland. But since the variety carried an 
appreciable percentage of off-type susceptible plants this could not account for the 
total absence of the NPA group from these States. Mid-summer temperatures in the 
linseed growing districts of these States exceeded the optimal, and possibly the 
maximum for development of the pathogen in the uredosoral state in the field. The 
combination of general host resistance and adverse summer temperatures probably 
accounted in large measure for the absence of the NPA groups. Summer temperatures 
in much of Victoria, South Australia, and Western Australia exceeded the optimal and 
possibly the maximum limits within which the NPA group could maintain itself 



56 PHYSIOLOGICAL SPECIALIZATION OF MELAMPSOKA LINI IN AUSTRALIA, 

vegetatively in the uredosoral stage. But the more southern sections where flax was 
most commonly grown had a sufficiently mild summer to permit the group to aestivate 
successfully if not prolifically in this stage. 

The occurrence of the PA group in all the flax- and linseed-growing parts of 
Australia, emphasized particularly by earlier race surveys, and also in districts where 
the commercial host was not grown or was restricted to very small experimental 
sowings presented one of the most interesting problems raised by the survey. Race 1, 
the only representative of this group identified during the current survey, was by far 
the most common race. It was the only Australian race found in all five States 
included in the survey. But it had the most restricted host range of all races 
identified during the survey. All commercial flax and linseed varieties were immune 
to it in the field. Such race 1-susceptible varieties as Punjab and Imperial were grown 
on far too small a scale and in too few localities for them to be considered the natural 
host of the PA group in the field. This suggested that the PA group did not maintain 
itself on the cultivated species, but rather on a widely distributed wild species. Such 
a host could be Linum marginale, native Australian flax. With the possible exception 
of parts of Queensland at a distance from the New South Wales border, this species 
was found in all flax- and linseed-growing districts of Australia, in many cooler 
highland districts bordering these districts, and at Tibooburra in inland New South 
Wales. 

It would be interesting to know to what extent the distribution of the PA group 
was limited by higher summer temperatures. The common occurrence of the group 
in New South Wales and Queensland, at such places as Tibooburra in the former 
State and Biloela in the latter State, showed clearly that race 1 was much more 
tolerant to high summer temperatures in the field than NPA races. It was able to 
aestivate vegetatively in the uredosoral stage at Sydney, although an NPA race was 
completely eliminated from the field during the summer. Similar observations were 
made at Werribee in Victoria. In the linseed-growing zones of north-west New South 
Wales and Queensland, where summer temperatures are considerably higher, the 
pathogen may retreat to the cooler highland districts, just as the pathogen retreats 
from the hot plains of India (Prasada). While the only positive evidence from 
Queensland showed that race 1 did not survive in the uredosoral stage at Toowoomba 
beyond mid-summer, the race may well survive in cooler microlocal zones within such 
upland districts. The possibility of its survival on the inland plains was by no means 
ruled out, in view of the presence of the race at Tibooburra. 

The mode of survival of the two race groups from year to year seemed to differ 
considerably. Within the NPA group there was little evidence that a given race 
maintained itself for long periods or spread to any great extent from its centre of 
origin. In view of the multiplicity of races during the current survey, and the marked 
change since earlier surveys, the sexual cycle must constitute a major phase in the 
history of this group. It seems unwise to assume that two collections of rust from 
different States or from the same State in different years were genetically identical 
and vegetatively related, even though they gave the same reaction on a given set of 
differential varieties. 

Exhaustive race surveys were carried out by Cruickshank (1952, 1956) in New 
Zealand over what, by comparison with Australian conditions, was a relatively restricted 
area. There was a considerable carry-over of races in successive years. Since genetical 
investigations of host resistance and studies of uredospore longevity were being carried 
on simultaneously with the race survey, and since the writer was dependent on the 
kindness of field workers from other States for the supply of infected material, the 
same detailed survey of flax- and linseed-growing districts could not be carried out to 
assess the degree of carry-over of races within the NPA complex from one year to 
another. But a comparison of results obtained in the survey under discussion with 
results obtained in previous surveys, particularly in Western Australia, showed clearly 
that there had been a considerable shift in the NPA complex within a period of ten 
to thirteen years. 



. BY H. B. KEEE. 57 

Ottawa-attacking races were restricted to a small section of Western Australia 
between 1949 and 1951. The first race, race 7, was identified in 1949, Another race, 
race 14, was found in 1950, and three new races, 15, 16 and 17, were isolated in 1951. 
One would judge from this evidence alone that the NPA complex was in a state of 
constant pathogenic flux. Race 14 was the first such race to be identified in another 
State, when it appeared in Victoria in 1952. It is impossible to determine whether the 
two occurrences of this race so separated in time and space stemmed from a common 
source or whether the race in Victoria developed in situ. But unless infected material 
was transferred between the two States it seems wisest to postulate independent origin. 

Although a given race within the NPA complex may have failed to survive from 
one year to the next, new virulence factors, especially those with a high positive 
survival value, were not eliminated. The new factor conditioning virulence on Ottawa 
770B soon became well established in the NPA complex of Western Australia even 
though the original race which carried it did not seem to have survived or spread. 
A factor or factors conditioning wide virulence on the heterogeneous variety Walsh 
appeared in race 11 in Victoria in 1949. The race did not seem to survive. Had it 
done so, specimens of such a potentially serious race would certainly have been 
forwarded from the district for analysis. But the new virulence factors evidently 
survived, though masked in less virulent races, eventually recombining in an important 
Walsh-attacking race reported from the same district in 1954. 

The evident importance of the sexual cycle in the propagation of the NPA complex 
indicates the need for a careful study of this phase in the field. While there would 
not appear to be a Statewide suppression of the uredosoral stage in Victoria during 
summer, there did appear to be complete recession of this stage in some districts. This 
should increase the importance of the teleutospore stage as a carry-over phase. All 
stages in the life cycle of M. lini have been recorded on the cultivated host in New 
Zealand, as well as on the indigenous wild species L. monogynum. An intensive survey 
of southern flax-growing districts in Australia should indicate a like situation. This 
situation is most likely to obtain in the vicinity of flax mills, where harvested flax 
is stacked prior to processing. These stacks, by bringing together infected host and 
possibly different races from several districts, by buffering the teleutospore material 
against excesses of high summer temperature and ensuring a constant supply of 
volunteer plants, must provide those conditions by which the pathogenic potential of 
the pathogen is brought together, recombined and carried over from season to season. 
A visit to several flax mills near Melbourne in 1947 indicated an abundant supply 
of heavily infected volunteer plants. 

It would be interesting to know to what extent the NPA race complex was affected 
by the growth cycle of the cultivated host. Mild temperatures and abundant supply 
of immature susceptible host ensured a maximum proliferation of this group during 
spring and early summer in Australia. High summer temperatures and lack of host 
following harvest resulted in an annual recession of the uredosoral stage, complete in 
some districts, partial in others according to local temperatures and supply of susceptible 
volunteer plants. Even if the recession were only partial and the volunteer plant 
population were entirely non-selective in its carry-over, it is highly improbable that the 
old balance of races would survive to the following year. The more severe the 
recession of the uredosoral stage in any year, the more prominent should be the role 
of the teleutospore as the carry-over phase, the more marked should be the change 
in the NPA complex in the following year. The diversity of races in the NPA complex 
and the marked change since the first survey could not, however, be attributed primarily 
to the Australian crop cycle. In New Zealand, with its much milder summer 
temperatures and spring sowings, M. lini maintained itself throughout the year in 
the vegetative uredosoral stage. The same multiplicity of races was found, though 
there was a considerable carry-over of old races in successive years. 

The history of the PA group differed strikingly from the NPA group. Only one 
PA race, race 1, was detected during the four years of this survey. Specimens of this 
race were received from Western Australia, South Australia, Victoria, New South 



58 PHYSIOLOGICAL SPECIALIZATION OF MELAMPSOEA LINI IN AUSTRALIA, 

Wales and Queensland. In the last two States it was located at such remote districts 
as Tibooburra and Biloela. Three Punjab-attacking races were identified in earlier 
surveys. They were differentiated by reactions on varieties not considered reliable 
for race differentiation in the current survey. Such differences as then existed were 
extremely fine and none of the three PA races had a range of virulence exceeding 
the least virulent of the NPA races. Such differences within the PA group as existed 
between 1940 and 1942 presumably still existed between 1948 and 1952, but could no 
longer be detected with the then used differential series and the currently recognized 
reaction levels of these varieties. Allowing for possible minor fluctuations of the 
nature indicated above there was no change in the pathogenicity of the PA complex 
from the first survey in 1940 to the completion of this survey in 1952. This contrasted 
with the NPA group within which fourteen of the eighteen races identified during 
this survey were recognizably different from races identiiied in earlier surveys. This 
suggested that the PA race group maintained itself entirely vegetatively in all parts of 
Australia or that the group, if it completed the sexual cycle, was pathogenically 
homozygous and homogeneous for those factors conditioning its avirulence on the 
eighteen immune members of the differential series. Both factors probably contributed 
to the uniformity of the PA group in time and space. Owing to the pressure of the 
several lines of investigation, race 1 teleutospores were not studied to determine the 
degree of homozygosity of this race. This could be a very useful study, which might 
throw much light on the unusual geographic and time patterning of this group. 

Paralleling the remarkable pathogenic uniformity of the PA group was an equally 
notable uniformity in high temperature tolerance. This feature was lacking from 
the NPA group (Kerr, 1958) and must have contributed substantially to the PA 
group's capacity to penetrate and maintain itself in New South Wales and Queensland 
in a northerly sweep into higher temperature zones far beyond the line of the NPA 
group. If this tolerance were conditioned by a number of factors operating in 
complement, and some or all heterozygous, it might have contributed to the stability 
of the PA group by militating against the survival of ofl types lacking this tolerance. 
This seems unlikely since such races could have survived in the more temperate zones 
of the southern States. The mode of inheritance of high temperature tolerance could 
perhaps be determined relatively simply by studying the germination of uredospores 
of different segregates on host extract at temperatures in the vicinity of 20^C. (Kerr, 
1958). It seems possible, assuming that the PA group completes the sexual cycle, that 
it is homozygous for those factors determining high-temperature tolerance, as well as 
those factors conditioning avirulence on the eighteen immune differential varieties. It 
would follow that mutations which might have resulted in the appearance of recessive 
factors conditioning virulence on some at least of these immune varieties have occurred 
very infrequently, if at all. With so many pathogenic factors involved it is surprising 
that no change in the pathogenicity of the PA group was registered over a period of 
twelve years. 

Whether or not the sexual cycle played a major part in the survival of the PA 
group, it was evident that race 1 survived summer quite readily in the field in the 
uredosoral stage in areas where the NPA complex was suppressed. Given a year-round 
supply of susceptible host, the PA group could maintain itself vegetatively without 
difficulty in the southern States and in New South Wales, at least as far north as 
Sydney. This should certainly have tended to stabilize the pathogenicity of the group 
and might account in large measure for the uniformity of the group in time and space. 

The manner of origin of new races is posed again by these studies, but no 
conclusive evidence can be brought forward to support either of the two major theories. 
In brief, these theories may be summarized as follows: (1) Mutations at loci 
conditioning virulence are relatively frequent. Potential new races, many with widened 
host range, are constantly being generated. (2) Mutations occur very infrequently. 
Any appreciable variation in virulence is the product of centuries of accumulation of 
mutant genes. Genes previously masked by dominant avirulent allels or present in 
too low a frequency to ensure detection, even in the most thorough survey, may 



BY H. B. KERR. 59 

suddenly manifest themselves long after their origin when a change in the host complex 
suddenly confers a high positive survival value on races homozygous for such virulence 
factors. 

Although no conclusive evidence for either theory can be drawn from the data 
presented in this paper, a line of investigations may be suggested which could lead 
to some definite decisions. The potential of such studies would depend to some extent 
on the validity of conclusions concerning the origin of the PA group. But even if such 
conclusions, to be dealt with shortly, are not valid, the study could yield useful 
information so long as the NPA and PA groups continue to coexist as non- 
interpenetrating systems. 

Delaying consideration of this for the present it should be noted that while the 
evidence from the PA group tends to suggest that mutations are very infrequent, the 
evidence from the NPA group could be interpreted in favour of the theory that new 
races are constantly being generated by mutation of old races. Any evidence for this 
must assume that the change cannot be accounted for by introduction of new factors 
from outside sources. This assumes rigid policing of quarantine laws. The need for 
rigid quarantine measures has been fully recognized by those in a position to import 
seed. Such quarantine measures have become increasingly strict during the period 
1940 to 1952, as flax and linseed assumed increasing economic importance. Prior to 
the war, when acreages were extremely low, lesser precautions seem to have been taken, 
and one outbreak at least of the pathogen in this period was attributed to contaminated 
imported seed. 

While no natural quarantine barrier exists between the eastern States or these 
States and South Australia, there is an extensive desert barrier between Western 
Australia and the more easterly States. For this reason it seems reasonable to assume 
that the occurrence of an Ottawa-attacking race for the first time in Victoria several 
years after the first appearance of such a race in Western Australia represents an 
independent development of this race group. The development of this race group 
followed the release of new, supposedly resistant varieties, carrying Ottawa resistance. 
These varieties were developed w^ithin Australia. Had the appearance of the Ottawa- 
attacking group coincided with large-scale importation of seed this might have indicated 
that the race group traced its origin to the country of origin of this seed. That this 
was not so lends weight to the assumption that the group developed in situ in Western 
Australia. That Ottawa-attacking races probably developed independently in two States 
separated by a major quarantine barrier suggests that the Ottawa virulence factor 
either took its source in both States from recent mutations or was already present 
in the race complex in sufficient frequency to ensure its manifestation within a very 
short time of a change in the host from an unfavourable to a favourable population. 

Since 1947 races avirulent on Abyssinian (races 2 and 4), avirulent on Williston 
Golden (race 17), and avirulent on Very Pale Blue Crimped (races 12 and 18) appeared 
in the Australian NPA complex for the first time. Previous evidence rules out the 
possibility that factors conditioning avirulence on these varieties were derived from 
the PA group. Genetic studies suggested that the Williston Golden factor conditioning 
resistance to PA race 1 differed from the factor conditioning resistance to NPA race 17. 
The new avirulence factors may have been introduced from overseas, though the 
introduction of three such factors in so short a time seems unlikely. They may have 
been present in the original race complex at a very low frequency and escaped 
detection. But since three factors were involved, since they could not have been masked 
by dominant allels, and since several hundred collections were analysed during earlier 
surveys this is by no means certain. The possibility of their origin by recent mutation 
cannot be ruled out. 

It remains to determine the manner of origin of the two race groups. Many of 
the races now present in Australia stemmed from introduced races. McAlpine, reporting 
the first record of rust in this country, concluded that M. lini had been introduced 
from overseas. A rust epidemic in Victoria was attributed to the same cause in 1936. 



,i60. , PHYSIOLOGICAL SPECIALIZATION OF MELAMPSORA LINI IN AUSTRALIA, 

Flor, in a personal communication, noted that the NPA races resembled those 
obtained by him by hybridization of North American and South American races. He 
suggested that races had been introduced from North America and Europe and had 
hybridized in Australia with more widely virulent races from South America. 

The PA group almost certainly derived from some other source. Up to 1952 no 
such race as race 1, avirulent on Bison, had been recorded in North or South America. 
The avirulence of the North and South American races on BoUey Golden and Newland 
was conditioned by two factors and one factor respectively. The avirulence of PA 
race 1 on these varieties was conditioned by four and two factors respectively. 
Punjab-attacking races were identified in New Zealand, but none was avirulent on Bison 
(Cruickshank, personal communication). New Ze'aland Punjab-attacking races 3 and 13 
(reclassified 10 and S respectively) (Cruickshank, 1956) were compared at Sydney with 
several Australian accessions of race 1. The New Zealand races were much less 
virulent on Punjab than the Australian accessions. Limmi marginale was not included 
as a source of rust during the New Zealand race surveys (Cruickshank, personal 
communication), although it is an important hold-over host for the pathogen in that 
country. There is no information about the race complex present on this species, but 
the absence of Australian Punjab-attacking race types from commercial crops growing 
in the same districts as the wild species is in line with the assumption that the 
Australian PA group does not occur in New Zealand. 

Indian races closely resembled the Australian PA group in their virulence on 
Bombay (carrying the same gene for resistance as Punjab) and avirulence on Bison 
and most of the differential varieties. If the Indian races possess the high temperature 
tolerance of the Australian PA group, and carry four factors and two factors 
respectively conditioning avirulence on Bolley Golden and Newland, the case for the 
derivation of the Australian PA group from India is strong. 

This, however, cannot rule out the possibility that the PA group is endemic to 
Australia. Had the PA group been introduced in the same manner as the NPA group 
it is surprising that the two groups did not merge their pathogenic potential. Probably 
accounting for this and supplying the most critical evidence was the almost exclusive 
association of the native Australian flax Linnm marginale with the PA group. Twelve 
of thirteen races collected from the wild species between 1940 and 1952 were Punjab- 
attacking races. 

Wild flax species are common hosts to M. lint in all countries in which the pathogen 
has been studied, L. monogynum in New Zealand, L. rigidum, L. leivisii. L. angusti- 
folium and L. sulcatum in America, as well as several species in Europe. It would be 
surprising if the native Australian species had only been host to M. llni since the 
settlement of the country. According to McAlpine, the pathogen was first identified on 
the cultivated host in 1889. Waterhouse obtained records of the pathogen on 
L. marginale at Como near Sydney in 1887. It seems reasonable to assume that the 
association of M. lini with the wild species predated the introduction of races from 
overseas. If this were so it is strange that the original race complex should be so 
repressed by introduced races that twelve of thirteen races isolated from the wild host 
between 1940 and 1952 were a race or races of Indian origin. The races isolated from 
the wild host may well have represented the original Australian M. lini complex. 
This was supported by the occurrence of race 1 on L. marginale at Tibooburra in the 
extreme north-west of New South Wales, a location so far removed from districts 
where the cultivated host was grown that derivation of the pathogen from the 
cultivated host seemed most unlikely in this instance. Identification of the pathogen 
at Pucuwan near Temora and at Curlewis added confirmation. Although Curlewis 
fell within areas where linseed was being grown commercially there was no known 
source of M. lini in the commercial crops. An exhaustive personal survey in the 
north-west of New South Wales, from Tamworth to Moree, failed to detect the 
slightest trace of rust in any of the commercial crops. Close checks by officers of 
the Department of Agriculture were equally fruitless. No commercial crops of flax 
or linseed were grown within 150 miles of Sydney, but a competent observer reported 



BY H. B. KERR. 61 

that the wild species was commonly infected with rust in the Sydney area. The 
presence of the pathogen in New South Wales was almost certainly independent of 
the cultivated host. 

The only evidence against these conclusions was the similarity of the Australian 
PA group and Indian races. This cannot be ignored, but it is by no means conclusive. 
It is not unreasonable to suppose that Australian and Indian races should be 
pathogenically similar. They might well trace back to a common source. 

The association of the NPA complex with the cultivated host, L. usitatissivvmn, was 
reflected in the accumulation of a wide range of virulence factors specific for resistance 
factors carried by the various varieties within the species. The almost total lack of 
these factors in the PA group (stressed even more by the genetic investigations of 
host resistance to race 1) could be postulated as further evidence that the PA group 
was adapted to the wild species rather than the cultivated species. The PA group, 
despite its apparent pathogenic homogeneity, may well have accumulated a range of as 
yet undetected virulence factors specific for resistance factors in L. marginale. Since 
some lines of marginale were resistant to race 1 there must be diversity within the 
species. Should a thorough study of the wild species and the races present on it 
confirm the above, the case for an original Punjab-attacking Melampsora lini complex 
would be strong. No detailed study of the wild species has yet been made in Australia. 
Various lines were received from different States during these studies. Attempts were 
made to hybridize several of these lines with different varieties of L. usitatissimum 
using both species as female parents. The two species were completely incompatible. 

If the wild species is the natural host of the PA group, this group should be able 
to maintain its distinctive identity as effectively in the future as it did between 1940 
and 1952. A thorough study of this race complex might then indicate the extent 
to which the pathogen may generate new mutant virulence factors. This is Impossible 
with the NPA complex where the possibility of introduction of new races from 
overseas can rarely be ruled out with complete certainty, and where the frequency 
of the individual virulence factors is controlled by not infrequent changes in the host 
complex. The presence of pathogenic factors in the PA complex specific for factors 
in the cultivated host could be used to determine the rate at which such factors are 
generated. Such factors, unless also fortuitously specific for or associated with some 
factor specific for resistance factors in the wild species, should be present at a 
frequency determined by the balance between the natural rates of generation and loss. 

These studies confirm the well-established fact that a change in the host complex 
is generally followed by a change in the pathogen race complex. New varieties were 
released to counter the susceptibility of the common flax varieties Concurrent and 
Liral Crown. Shortly after the release of Wada and Boyup, Ottawa-attacking races 
appeared for the first time in Australia and soon dominated the Western Australia 
NPA complex. This stresses the need for combined resistance already emphasized by 
Watson and Singh (1952). 

The range of combined resistance in Linum usitatisshnum is greatly reduced by 
the limited number of loci in the host determining resistance to Melampsora lini. 
But this may be offset to some extent by the constant pathogenic flux of the NPA 
group. Flor (1946) established a one-to-one correlation between resistance factors in 
the host and specific virulence factors in the pathogen. Avirulence was always 
dominant to virulence. The greater the number of resistance factors in a host 
variety, the less likely is the pathogen to accumulate the requisite virulence factors 
homozygously in the one race. If the pathogen survives via the teleutospore stage 
rather than vegetatively from one season to another, the survival of a given complex 
of virulence factors becomes increasingly uncertain in a mixed race situation the 
greater the number of virulence factors involved. 

During these studies, Wada, with its resistance derived from a single Ottawa 
factor, succumbed to an Ottawa-attacking race soon after its release for commercial 
production. While the original race did not survive, the Ottawa-attacking virulence 



bti PHYSIOLOGICAL SPECIALIZATION OF MELAMPSOKA LINI IN AUSTRALIA, 

factor carried over in the race complex from one year to the next in Western Australia. 
Walsh, by contrast, the only commercially grown linseed variety, grown for a con- 
siderable period over a wide geographic range, remained effectively immune or resistant 
to the Australian PA and NPA race groups. One Walsh-attacking race appeared in 
one district in Victoria, but did not spread or survive to the next year. Genetic 
studies have shown that Walsh, though heterozygous, owes its resistance to at least 
three factors. 

Acknowledgements. 

This work was carried out at the Faculty of Agriculture, Sydney University, first 
under a grant from Meggitt Ltd. and later during tenure of a Thomas Lawrance Pawlett 
Scholarship. Particular thanks are due to Professor W. L. Waterhouse for his initial 
direction of the work and his constant encouragement, and also to Associate Professor 
I. A. Watson and other members of the Faculty for advice and assistance given during 
these studies. 

References. 
Cass SmitHj "W. P., and Harvbt, H. L., 1946. — Flax rust in Western Australia. Jour. Dept. 

Agr. Western Australia, 23 : 42. 
Charles, A. W., 1947. — Honours Thesis, Faculty of Agriculture, Sydney University. 
Cruikshank^ I. A. M., 1952. — Studies on the Physiologic Specialization of Melamp.sora lini 

(Ehrenb.) Lev. in New Zealand. N.Z. Jour. Sci. and Tech., Section B, Vol. 34, No. 2. 
, 19 56.- — A Further Note on the Physiologic Specialization in Melampsora Hni (Ehrenb.) 

Lev. in New^ Zealand. N.Z. Jour. Sci. and Tech., Section B, Vol. 38, No. 2. 
Deerbtt p. H., 1954. — Linseed Notes. The Australian Plant Breeding and Genetics News- 
letter, No. 4, 1954, p. 9. 
Flor, H. H., 1935. — Physiologic Specialisation of Melampsora lini on Linum usitatissimum. 

Jour. Agr. Res., 51 : 819-837. 
, 1945. — Analytical Key for the Identification of Physiologic Races of Melampsora lini. 

U.S.D.A. Mimeographed Bulletin, Nth. Dak., 20 pp. 
, 1946. — Genetics of pathogenicity in Melampsora lini. Jour. Agr. Res., 73: 335-357. 



FoRSTER, H. C, 1941. — The Use of Climatic Graphs in Determining Suitable Areas for Flax 

Production. Jour. Vic. Dept. Agr., 39: 515-524. 
Gray, S. G., 1950. — Variety Trials with Linseed in Southern Queensland: A Progress Report. 

C.S.I.R.O. Div. Plant Indust., Divisional Rept. No. 8. 
Kerr, H. B., 1951. — The Use of Excised Shoots in Linseed Investigations. Proc. IjInn. Soc. 

N.S.W., 76: 183-186. 
Kerr, H. B., 1952. — Rust Investigations in Linseed. Nature, 169 : 159. 
, 1958. — Melampsora lini (Pers.) Lev. Uredospore Longevity and Germination. Proc. 

Linn. Soc. N.S.W., 83 : 259-287. 
McAlpine, D., 1906. — The Rusts of Australia. 349 pp. 
MiLLiKAN, C. R., 1951. — Diseases of Flax and Linseed. Vic. Dep. Agr. Tech. Bull. No. 9, 

140 pp. 
Prasada, R., 1948. — Studies in Linseed Rust, Melampsora lini (Pers.) Lev. in India. Indian 

Phytopath., 1 : 1-18. 
Straib, W., 1939. — Untersuchungen iiber den Wirtsbereich und die Aggressivitat physiologischer 

Rassen von Melampsora lini (Pers.) Lev. Zuchter., 11: 130-136, 162-168. 
Thomas, I., and Millington, A. J., 1946. — Flax and Linseec? Breeding in Western Australia. 

Wada, a New Rust Resistant Flax Variety. Jour. Dep. Agr. West. Auat., Series 2, 

23: 39-42. 
VallegAj J., 1944. — Especiallzacion Fisiologica de Melampsora lini en Argentina. Santa 

Catalina Inst. Fitolec, Pub. 39. (De los Anales del Instituto Fitotecnico de Santa Catalina, 

4 (1942) : 59-74.) 
Waterhouse, W. L., and Watson, I. A., 1941. — Note on Physiological Specialisation in Flax 

Rust. Proc. Linn. Soc. N.S.W., 75: 115-117. 
, 1943. — Further Determinations of Specialisation in Flax Rust Caused by Melampsora 

lini (Pers.) Lev. Proc. Linn. Soc. N.S.W., 77: 138-144. 
Watson, I. A., and Singh, D., 1952. — The Future for Rust-Resistant Wheat in Australia. 

Jour. Aust. Inst. Agr. Sci., 18 : 190-7. 

EXPLANATION OF PLATES III-IV. 
Plate ill. 
1. — Fully susceptible type 4 reaction. Compound pustules with no chlorosis. Leaves 
rarely distorted. Note small newly developing non-compound pustules on same leaf. x 3^. 
2. — Typical fully susceptible reaction without compound pustules. No chlorosis and 
infection general over whole leaf surface, x 3J. 



BY H. B. KERR. 63 

3. — Fleck type immune reaction. Small chlorotic flecks over most of leaf surface, x 3J. 

4, 5, 6. — Various intermediate resistant reactions with pustules of varying size and' varying 
degrees of chlorosis and necrosis. Infection often localized to part of the leaf, and leaf 
commonly misshapen, x 3J. 

Plate iv. 

1. — Bulk inoculation of F257 with race 1. Fairly heavy production of uredosori on leaves, 
but very few developing on stems. Plants recovered and still growing vigorously, x 1. 

2. — Two sets of excised shoots of Concurrent inoculated and incubated at the same time 
under identical conditions with an Ottawa-attacking- race, race 17 (LHS), and a non-Ottawa- 
attacking race, race 6 (RHS), Note complete recovery of shoots inoculated with race 17 
and the clean uninfected stems. Shoots inoculated with race 6 developed heavy stem infection 
and failed to resume noiinal growth, x 3,. 

3. — Bulk inoculation of Punjab with race 1. Very heavy development of uredosori on 
leaves and stem. Growth of shoots severely retarded, x i. 



64 



THE FAMILIES OF CYCADS AND THE ZAMIACEAE OF AUSTRALIA. 
By L. A. S. Johnson, National Herbarium, Royal Botanic Gardens, Sydney. 

(Four Text-flgures.) 

[Read 25th March, 1959.] 



Synopsis. 

The general classification of the Cycadales is reviewed and grounds are put forward for 
the recognition of three families: Cycad'aceae (Cycas), Stangeriaceae, fam. nov. (,Stangeria) 
and Zamiaceae (remaining- genera). A taxonomic revision of the Australian members of 
the Zamiaceae is provided. Three genera, all endemic, are recognized' : Lepidozamia Regel 
(2 spp. ), Macrozamia Miq. (14 spp. : 6 in sect. Macrozamia, 8 in sect. Parazamia) , and 
Boivenia Hook, ex Hook. f. (2 spp.). New taxa are described and new combinations made 
as follows: Lepidozamia hopeites (Ccokson) L. Johnson (fossil species), Macrozamia 
communis, M. diplom,era (F. Muell.) L. Johnson, M. lucida, M. stenom,era, M. pauK-guilielmi 
W. Hill and F. Muell. sspp. plvirinervia and flexuosa (C. Moore) L. Johnson. The name 
M. spiralis (Salisb.) Miq. is shown to be correctly applied to the species known as M. corallipes 
Hook. f. The species known as M. spiralis in Queensland and New South "Wales are the 
n. spp. M. lucida and M. communis respectively. Keys and discussions are provided for all 
taxa and the very confused synonymy is reviewed and' clarified. 



Introductory. 
This study has arisen out of the necessary revision of the New South Wales species 
of Macrozaynia for the forthcoming Flora of Neio South Wales, part 1 (in press). It 
was clear from the outset that the specific limits needed clarification and that the 
nomenclature was in a state of chaos. Moreover, it soon became evident that generic 
as well as specific concepts were at issue and this in turn led to a consideration of 
the general taxonomy of the cycads. 

Part I. The Families of Cycads. 
Ge7ieral. 

It has been customary to refer all the true living cycads (universally accepted as 
the order Cycadales) to a single family, Cycadaceae, variously divided by different 
authors into subfamilies, tribes and subtribes. Schuster (1932, p. 63) gives a synopsis 
of these arrangements. Amongst these categories, however, all authors have recognized 
a suprageneric taxon including Cycas alone. In recent times only Wettstein (1923) 
has placed Cycas in a family (Cycadaceae sensu stricto) by itself, grouping the 
remaining genera as Zamiaceae. The latter family had, however, been established 
much earlier by Reichenbach (1837), though several cycad genera were then unknown. 

Taxonomists have a tendency to recognize very inclusive families in the more 
unfamiliar groups of plants. The cycads are rather few, they are all very different 
from other living plants and they are clearly related to each other; consequently 
most botanists, impressed by this apartness, are content to lump them together and 
to minimize the differences within the group. Now all taxonomic classification, at 
least above the specific level, is to a considerable degree subjective as regards the 
status of admittedly related taxa. To take familiar examples from the flowering plants, 
the dismembering of the Leguminosae s. lai. into the three families Mimosaceae, 
Caesalpiniaceae and Papilionaceae must be more subjectively based than the exclusion 
of Paeoniaceae from Ranunculaceae. In the former case few will deny that the 
segregate groups have a closer phylogenetic relationship among themselves than any 
one of them has to other living families; the claim for family status rests on the 

Proceedings of the Linnean Society of New South Wales, 1959, Vol. Ixxxiv, Part 1. 



BY L. A. S. JOHNSON. 65 

rather indefinable (though not unreal) degree of difference between, and coherence 
within, the groups. In the latter case morphological, anatomical and cytological 
grounds exist (Eames, 1953; Cronquist, 1957) for believing that Faeonia is in fact 
closely allied to certain families (the Dilleniales line) quite different from 
Ranunculaceae, and that it has been placed in the latter family on the basis of 
superficial resemblance; this is an objectively based segregation, provided that the 
general system of families in the dicotyledons is accepted as a frame of reference. 

The small, self-contained group of the cycads lacks any such established frame of 
reference and the case for segregate families is thus scarcely a provable one. It is 
nevertheless a reasonable one. We can show that certain genera have many characters 
in common and could easily be derived from a common ancestral population which 
(at this stage) would still be quite different from the conceivable ancestors of other 
living genera. There are, furthermore, no links or breakdowns between these groups. 
Thus the groups are natural and their distinctions indicate evolutionary divergence 
at an early stage in the history of the Order. To justify family status we can apply 
to no fixed criterion, but can say that the differences are of such a nature that common 
ancestry for the whole assemblage must be very remote and that each of the family 
groups must be the result of considerable evolution since the divergence, the present 
members doubtless being relict forms with many lost relatives. Unfortunately, the 
fossil record of the true cycads is scanty (see, for instance, Cookson, 1953) in contrast 
to those of their distant relatives the Mesozoic Bennettitales and the broad, more or 
less ancestral, chiefly late Palaeozoic group of the Pteridosperms, and it is not at 
present likely to throw much light on evolution within the Cycadales. 

On the principles outlined above, the Cycadales may be classified in three families: 
Cycadaceae s. str., Stangeriaceae and Zamiaceae, the last of which may be subdivided 
(with less assurance) into tribes. The genera as commonly recognized are entirely 
natural with the exception of Macrozamia, from which Lepidozamia must be excluded 
(see below, p. 83). 

There is little point in detailing the history of the many previous classifications, 
but of the more recent systems three may be mentioned. Hutchinson (1924) recognized 
two tribes, Cycadeae (Cycas) and Encephalarteae, the latter comprising the following 
subtribes: Encephalartinae* (Dioon, Macrozamia, Encephalartos) , Stangeriinae* 
{8tange7-ia) , Zamiinae* (Bowenia, Geratozamia, Microcycas, Zamia) ; this is a reason- 
able arrangement and comes close to that adopted here apart from differences in 
status and Hutchinson's curious placing of the distinctive Stangeriinae between his 
other two more similar subtribes. I have used Hutchinson's key to some extent in 
devising the one hereunder. Pilgor (1926) recognized five subfamilies, of which the 
first four were monotypic, containing Cycas, Stangeria, Boioenia and Dioon respectively; 
the fifth subfamily (Zamioideae) included the remaining genera without subdivision. 
The only merit of this is the recognition of Stangeria as rather distinctive. The 
subfamilial segregation of Boivenia and Dioon, while retaining such genera as Zamia 
and Macrozamia in a single equivalent taxon, is a characteristically unrealistic 
consequence of the analytical "key-character" used as a basis of classification. 
Schuster (1932), the latest monographer of the Cycadales, has two subfamilies, 
Cycadoideae (Cycas) and Zamioideae, the latter comprising eight tribes, each containing 
but a single genus! He does not say what purpose this is meant to serve. 

The present system, unlike those of Pilger and Schuster, is derived not by the 
use of key-characters, but synthetically by marshalling like genera together using the 
whole complex of their characters, and analytically by the recognition of fundamental 
and irreconcilable differences. Parallelism and convergence are pitfalls, as usual, and 
grouping within the Zamiaceae is difficult. As to characters, the three very different 
systems of leafiet venation are surely of greater evolutionary significance and more 
difficult to derive from each other than are some of the different conditions in 

* Given here in tlie correct forms under tlie present International Code of Botanical 
Nomenclature (Lanjouw, 1956) ; Hutchinson wrote "Encephalartineae", "Stangerineae" and 
"Zamineae". 

E 



6b THE FAMILIES OF CYCADS AND THE ZAMIACEAE OF AUSTRALIA, 

reproductive characters traditionally considered of more importance. When critically 
examined, these latter, with the exception of those distinguishing Cycas, are seen to 
be nothing more than modifications of the shape and regularity of arrangement of 
the essentially similar sporophylls. 

The chromosome numbers (from Darlington and Wylie, 1955) are not particularly 
enlightening in the cycads. They support the distinction of Cycas (x = 11, 12?), but 
in the Zamiaceae, while a; = 9 is found in five genera and o; = 8 in two (Zamia has both 
numbers), one genus (Microcycas) has x = 13. Microcycas agrees morphologically with 
the other Zamiaceae, but its chromosome number at least suggests a long history as an 
independent genus. The female gametophyte of Microcycas is remarkable for the very 
large number of archegonia and the male for the large number of spermatozoids 
produced (Chamberlain, 1919). The genus has no resemblance to Cycas, and certainly 
does not seem to be generally primitive in the Zamiaceae. Its chromosome number 
presents an intriguing problem, assuming, of course, that the count is correct. Finally, 
Stangeria has x - S but is very different from the two Zamiaceae (Zamia and 
Ceratozamia) with this number. The course of caryotype evolution is obscure. 

The characters of the Cycadales as a whole may be found in any of the standard 
works (e.g. Schuster, 1932); The Living Cycads (Chamberlain, 1919), though semi- 
popular in style, still provides an excellent extended account and comparative discussion. 

Key to the Families and Genera.* 

CYCADALES. 

1. Pinnae with a single thick midrib and no lateral veins, circinately involute in bud. Female 
sporopliylls not forming a determinate cone but spirally arranged in a terminal mass, and 
falling separately at maturity, the central axis eventually continuing vegetative growth 
(male sporophylls in definite cones). Female sporophylls with a pmnatifid, pectinate or 
toothed "lamina", ovules 2-several, marginally inserted proximal to the lamina, obliquely 
directed outwards ("ascending"). Trunk clothed with old frond-bases, x = 11, 12? Old 

World tropics I. Ctcadacbae. 

1. Cycas 

1.* Pinnae (or pinnules) with many lateral or longitudinal nerves, usually straight in bud 

(frond' circinate as a whole in Ceratozamia). Sporophylls of both sexes In determinate 

cones, the female sporophylls scale-like, more or less peltate with a thickened and laterally 

expanded end on the axis-facing margins of which the 2 (sometimes 3 in Lepidozamia and 

rarely 3 or more in other genera) inward'- facing ("inverted") ovules are inserted. Caudex 

various. 

2. Pinnae penniveined, with a definite midrib and numerous transverse parallel dichotomously- 

branched lateral veins, convolute in bud. Sporophylls imbricate but In almost vertical 

rows. Caudex sxibterranean, naked (frond-bases deciduous). Superficially fern-like 

plants, a; = 8. E. and S.E. Africa II. Stangeriaceae. 

2. Stangeria. 

2.* Pinnae (or pinnules) lacking a midrib, with numerous more or less parallel longitudinal 

nerves (dichotomously-branched near the base), imbricate but not convolute in bud. 

Sporophylls and caudex various, x = 8, d, 13 III. Zamiaceae. 

3. Sporophylls imbricate, not in vertical rows. Caudex clothed with persistent frond- and 

cataphyll-bases. Pinnae not obviously articulate, though somewhat deciduous when 

old in a few species. 

4. Cones axillary (sometimes erect and appearing falsely terminal, but not terminating 

ma,in axis). Ovules sessile on the sporophyll. Female sporophylls greatly thickened 

towards the ends and tightly Imbricate, glabrous or tomentose but scarcely woolly. 

a; = 9. Australia and Africa a. Tribe Encephalarteae. 

5. Sporophyll-ends acutely or bluntly pointed or with a terminal spine (sometimes 
reduced to a narrow transverse wing, but never a facet). Cones sessile or stalked. 
Australia. 
6. Cones sessile or subsessile ; sporojjhyll-ends tomentose, produced into a spreading 
obtuse to acute but not spinescent wing, curved upward or downward. Succes- 
sive crowns of fronds markedly separated by broad series of cataphylls. Pinnae 
inserted on the adaxial midline of the rhachis. Frond-bases shortly tomentose. 
E. Australia 3. Lepidozamia. 

* The terminology is the same as that used in Part II (see p. 77). Hutchinson's (1924) 
key and the excellent vegetative key of Kegel (1876) have been used freely in constructing- 
this one which, however, is based as far as possible on actual material. The key does not, 
of course, give all the distinguishing characteristics of the taxa. Each genus has a characteristic 
facies due to the form and arrangement of the fronds and their segments. 



BY L. A. S. JOHNSON. 67 

6.* Cones stalked ("pedunculate"); sporophyll-ends glabrate, often glaucous, at 
least the females subterminally compressed to form a more or less vertical 
surface, on which (in both sexes) the margins form a transverse ridge 
terminating in a rigid more or less erect spine. Successive crowns of fronds 
little separated by a few cataphylls. Pinnae inserted near the edges of the 

rhachis. Frond-bases silky or woolly with long hairs, a; = 9. Australia 

i. . . 4. Macro^amiav 

5.* Sporophyll-onds truncate, more or less prismoidal, with a 4- or G-sided terminal 

facet, somewhat decurved. Cones sessile, a; = 9. Africa 5. Encephalartos. 

4.* Cones (sessile or shortly stalked) terminating growth of the main stem, which 
continues vegetative growth sympodially from the base of the cone (the subterminal 
"cone-dome" ultimately engulfed by the new growth so that the trunk appears 
continuous). Ovules u.sually borne on a short stalk-Iike outgrowth (false funicle) 
of the sporophyll. Female sporophylls somewhat thickened but loosely imbricate 
at least at the ends (which are reduced in D. spinulosum) , the woolly end's flattened 

and erect, a; = 9. Central America 6. Tribe Dioeae. 

6. Dioon. 

3.* Sporophylls apparently valvate, arranged in vertical rows, their ends (at least in the 

females) hexagonal or rhomboid. Caudex various, often naked. Cones terminal 

(always?) in origin, on the main stem or short branches, though often pushed aside 

by new stem growth. Pinnae definitely articulate at the base or, if not, then fronds 

decompound c. Tribe Zamieae. 

7. Sporophylls either shortly woolly or 2-horned at the end. Trunk robust, clothed 
with leaf-bases and cataphylls or at length naked. Fronds simply pinnate, pinnae 
articulate. 

8. Sporophyll-ends obtuse or truncate, not horned, tomentose with short, crisped hairs. 

Frond-bases and cataphylls said to be at length deciduous, a; = 13. Cuba 

7. Microcycas. 

8.* Sporophyll-end's 2-horned, not tomentose. Frond-bases and cataphylls persistent. 

a? = 8. Mexico 8. Ceratozamia. 

7.* Sporophylls truncate, not woolly or horned. Caudex naked, usually slender or short, 
sometimes subterranean. Fronds either decompound or, if simply pinnate, then the 
pinnae articulate. 

9. Fronds simply pinnate, pinnae articulate at the base, a; = 8, 9. Tropical N. and S. 
America 9. Zamia. 

9.* Fronds decompound, pinnae and pinnules not articulate at the base, a; = 9. N.E. 
Australia 10. Bowenia. 

Enumeration and Discussion. 

In this paper I can neither list the detailed synonymy nor deal at length with 
the extra-Australian taxa, other than those newly established herein. Those interested 
may consult the monograph of Schuster (1932) for generic descriptions and biblio- 
graphy, but will be wise not to rely on it, especially at the specific level. Probably all 
the larger genera need thorough revision, en modern lines, by workers familiar with 
most of their species in the field. 

I. Family CYCADACEAE L. C. Rich.* in Pers., Enchir., 2 (1807), 630. 

As here restricted this includes only the genus Cycas. This is the most distinct 
of all the genera of cycads and in many respects preserves the most primitive 
characters, especially in the frond-like megasporophylls and their loose undifferentiated 
arrangement. However, Cycas cannot be considered to represent a form ancestral to 
the other living genera, since its single-veined pinnae could hardly have given rise to 
either the Stangeriaceous or the Zamiaceous condition. 

The circinate vernation of the pinnae (not the whole frond) is an apparently 
primitive feature. The much greater specialization of the male as compared with 
the female structures is the most remarkable feature of the family. This indicates 
that determinate male and female cycad "cones" may not be homologous as complete 
structures, but only in so far as they are aggregations (attained at different stages) 
of sporangium-bearing reduced determinate branch-systems (sporophylls) more or less 

* As "Cycadeae" but in family rank. 



€8 THE FAMILIES OF CYCADS AND THE ZAMIACEAE OF AUSTRALIA, 

equivalent to the fronds which (phylogenetically speaking) are somewhat less reduced, 
similar but sterile branch-systems. 

1. CrcAS L. Type species: C. circinalis L. . ^ 

Perhaps 20 species, extending to Madagascar and East Africa but chiefly in 
South-East Asia, tropical Australia and the western Pacific. Chromosome numbers: 
X - 11, 12? (2n = 22, but 24 in one count). 

Schuster (1932) recognizes only eight species but his treatment would appear to 
be no sounder than that of Macrozamia (see below, p. 72). Cycas is much in need of 
satisfactory revision but this would certainly require a wide knowledge of the species 
in the field and extensive living and herbarium collections from the whole of its wide 
range. Much herbarium material is practically useless. For these reasons I have not 
attempted to deal with the Australian species. The synonymy also is most complex 
and confused, as in Macrozamia. A diificult and protracted task awaits any responsible 
monographer of the genus. 

II. Family STANGERLACEAE L. Johnson, fam. nov. 

I can trace no previous publication of this taxon in family rank. 

Familia inter Cycadalibus frondium pinnls unicostatis arete penninerviis (venis 
dichotomis rectiusculis) vernatione rectis sed convolutis* distincta. Sporophylla 
utriusque sexus conos determinates formantia, squamiformia peltataque, mascula 
spor^giis multis infra instructa, feminea ovulas dua? versus axem coni directas in 
marginibus gerentia. Chromosomata ut .a; = 8 numerata. 

Genus typicum (unicum): Stangeria T. Moore. 

Stangeria has not usually been considered as distinct as Cycas from the other 
cycad genera, and indeed there is no apparent character in the reproductive structures 
Inconsistent with its inclusion in the Zamiaceae. However, the view that "good" taxa 
above the rank of species (or even species in some opinions) must always differ in 
reproductive characters is surely an example of archaic formalism lacking any firm 
evolutionary basis. "Good taxonomic characters" have no absolute significance and can 
be evaluated only in relation to the whole complex of circumstances in particular cases. 
A "conservative" character in one series of organisms may be quite radically modified in 
a closely related series or even in individual species within the same series. 
Evolutionary processes are no respecters of "characters", and useful as generalizations 
about characters may be when discussing trends, to apply them dogmatically in 
classification is to reduce taxonomy to a formal game rather than a branch of biology. 
This is in contrast to the view of Davis (1952, p. 149). 

It is clear that the strongly-developed midrib and the many crowded, spreading, 
lateral veins of Stangeria pinnae must develop by organogenetic growth processes 
considerably different from those which produce the longitudinally-nerved midribless 
frond segments of the Zamiaceae. Neither condition is easily derivable from the 
other. t This implies an early separate development along two lines from some 
ancestral group with a more generalized vascularization of the fronds. 

Anyone familiar with living plants of all or most of the cycad genera must be 
struck by the divergent appearance of Stangeria. Its fern-like, rather Marattiaceous, 
aspect is quite unlike that of any of the Zamiaceae or of Cycas. The resemblance to 
a Marattia or an Angiopteris is, of course, due to convergent, or perhaps more properly 
to parallel, evolution of a similar frond pattern from the more primitive forms of 
megaphyllous leaf (major determinate branch system) found in the very early 

* The vernation of the pinnae is usually described (e.g. by Schuster, 1932) as plicate; 
in the living material which I have examined the young- pinnae are strongly convolute and 
certainly not plicate, except perhaps at the tips. 

t It is conceivable that the simply pinnate frond of Stangeria is equivalent to a bipinnate 
type of Zamiaceous frond (cf. Bowenia) in which individual pinnules have underg-one 
"phylogenetic fusion". However, this seems very doubtful and Stangeria could still not be 
■derived from anything like the known Zamiaceous g-enei-a — certainly not from Bowenia. 



BY L. A. S. JOHNSON. 69 

Pteropsida. There is no particular relationship between the groups, which differ 
profoundly in reproductive structures and internal anatomy. On the other hand, the 
difference between Stangeria and the Zamiaceae is certainly not superficial, and 
suggests that some of their reproductive resemblances may be due to parallel 
development from early proto-cycadalean ancestors. With little doubt the naked 
tuberous stem and the geophytic habit are secondary and do not imply any relationship 
between Stangeria on the one hand and Bowenia or Zamia on the other. 

Such internal schisms notwithstanding, the unity of the Cycadales as a whole 
seems clear. The correspondence of all three families in structure and organization is 
too close to be the result of convergence from unlike ancestors, though parallel 
development of different lines from a common ancestral group seems probable. For a 
comparative discussion of homology, parallelism and convergence, and their significance 
in classification, see Simpson (1945, pp. 8-12). In the present paper I have attempted 
to apply principles of evolution-reflecting classification developed from considerations 
similar to those so admirably expressed by Simpson in this remarkable tou7- cle force. 

2. Stangeria T. Moore. Type species: S. eriopus (Kunze) Nash (&'. paracloxa 
T. Moore). 

A single variable species (as usually treated) in south-east Africa. Chromosome 
number: x - '8. (2n = 16). 

III. Family ZAMIACEAE Reichenb., Hanclb. (1837), 139. 

Eight genera and perhaps 80 species in the tropical to warm-temperate regions of 
Africa, Australia and North and South America. 

The Zamiaceae appear to be a natural group. All the genera certainly agree in 
many fundamentals of structure despite considerable diversity of detail and size. The 
characteristic longitudinal dichotomous venation has already been stressed when dis- 
cussing the other two families. The basic chromosome number .t = 9 predominates, with 
3" =: 8 (a reduction? Not found in the apparently more primitive genera) in two geneia 
and the anomalous-seeming a; = 13 in Microcycas. Chamberlain (1919) indicates various 
reduction and specialization series from genus to genus, but these are by no means 
always concurrent in different organs and it is obvious that, although some of them have 
a greater number of advanced features than others, the living genera do not represent 
an evolutionary series but rather the present end-points of a number of different lines, 
doubtless cognate with others now extinct. To judge from the phylogenetic series 
which many authors put foi'ward even today for Recent organisms of all kinds, it is 
still necessary to state this evolutionary platitude. 

With some diffidence I have attempted to arrange the genera into three tribes. 
These are perhaps not entirely natural (especially Zamieae, which includes the rather 
non-conforming Microcycas) , but they do bring together similar genera, some of which 
are undoubtedly not too distantly related. 

Phytogeographers should beware of basing any far-reaching arguments upon the 
distributions of these rather tentatively defined tribes. 

a. Tribe Encepiialaeteae (Miq. ) L. Johnson, comb, nov.* 

Three genera, in Australia and Africa. 

These genera have relatively few advanced features, though certain specializations 
are found. Although each genus is very distinct, a true relationship seems likely. 
F. Mueller at various times united them under EncejJhalartos and indeed suggested 
sinking this under Zamia; he did not believe in evolution (see below, p. — ), but in the 
individual creation of species and the consequent artificiality of genera. Affinity, apart 
from mystical connotations, meant merely resemblance in certain chosen characters. 

* Cycadaceae ("Cycadeae") tribe Encephalarteae Miq., Frodr. Syst. Cycad. (1861), 5. 
Schuster (1932, p. 64) attributes this to himself, in .spite of his citation on p. 63 of its use- 
by Miquel and other authors. 



70 ■ THE FAMILIES OF CYCADS AND THE ZAMIACEAE OF AUSTRALIA, 

3. Lepidozamia Regel. Type species: L. per-offskyana Regel. 

Two species in tropical and subtropical eastern Australia. Chromosome number 
unknown. 

Discussed in detail in Part II (p. 83). 

4. Maceozamia Miq. Type species: M. riedlei (Fisch. ex Gaudich.) C. .A.. Gardn. 
(see p. 87). 

Fourteen species in subtropical and warm-temperate Australia. Chromosome 
number: x = 9 (2n = 18). 

Discussed in detail in Part II (p. 87). 

5. Encepiialartos Lehm. Type species to be selected by a monographer. 
Fit'teen-twenty species in Central and South Africa. Chromosome number: x - d 

(2w = 18). 

The genus still needs some revision, but the tropical species have been fully treated 
by Melville (1957). 

b. Tribe Dioeae (Schuster) L. Johnson, comb, nov.* 

A single genus in Central America. 

This tribe was rather meaningless as set up by Schuster, since all his tribes were 
unigeneric. However, Dioon does stand somewhat apart from the Encephalarteae, their 
common characters being chiefly merely the generally primitive ones of the family. It 
may be convenient to allot it a tribe to itself. The cones are terminal (according to 
Chamberlain's studies), as in the Zamieae, but Dioon differs considerably from that 
tribe in other characters. 

6. Dioon Lindl. Type species: D. edule Lindl. 

Three (or more?) species in Central America (chiefly Mexico). Chromosome 
number: x - 9 (2w = 18). 

Dioon and Lepidozamia have the least modified megasporophylls in the Zamiaceae. 
They are otherwise not particularly alike. 

c. Tribe Zamieae.! 

Four genera, in America and .Australia. 

As state dabove, this tribe possibly includes the products of convergence rather than 
close affinity. The genera are grouped mainly on the regular arrangement of the 
sporophylls in the cone, in which definitely vertical orthostichies are apparent. The 
sporophylls are, of course, at the same time arranged in parastichies (spirals) as in 
(the other tribes. Phyllotactic modification to the regular vertical arrangement could 
easily have taken place more than once. Bowenia is geographically isolated from the 
other genera and its resemblance to Zamia in sporophylls, cones and habit could be 
secondary, at least in part. Microcycas has 13 gametic chromosomes and some gameto- 
phytic peculiarities, and thus stands rather apart from the other genera. Ceratozamia 
shares the number a- = 8 with some species of Zamia, but this probably does not indicate 
any special relationship. In habit, at least, Ceratozamia seems more primitive than 
Zamia and its sporophylls are distinctive. These American genera and Boivenia need 

* Cycadaceae tribe Dioeae ("Diooneae") Scliuster, Pflanzenr., IV, i (1932), 64. Schuster's 
spelling is to be corrected, as above. Since Dioon is from the Greek 5i and wov, and the stem 
of the latter, transliterated, is "5-" (cf. "oospore", "Oidium"), then the stem of the compound 
is surely "Dio-". Under the I.C.B.N. the tribal ending "-eae" must be added to the stem. 
Classical Latin authors, supposing- them to have used such a word at all, would doubtless 
have latinized it to the less outlandish-looking "Dioveae". One may perhaps enter a protest 
against the too common pronunciation of Dioon to rhyme with "soon" ; it rhymes with "so on". 
Lindley's original publication as '•Dion" was a mere slip, later corrected. 

t As the tribe containing the type, this requires no author citation. It was first established 
as Cycad'aceae ("Cycadeae") tribe Zamieae Reichenb., Consp. Regn. Veg. (1828), 40. (Not 
seen; the reference is from Schuster (1932, p. 63). On p. 64 he cites his own name as author, 
though Zamieae had been used by many earlier workers.) 



BY L. A. S. JOHNSON. VI 

a comparative study, with taxonomic rather than purely morphological principles in 
mind. Whether the cones are in fact always terminal in origin, as stated by Chamberlain 
(1919), needs to be confirmed. The cones of Ceratozamia, at least, appear lateral to the 
external view. 

7. MiCROCYCAS (Miq. ) A. DC. Type species: M. calocoma (Miq.) .A.. DC. 
A single species in Cuba. Chromosome number a; = 13 (2n = 26). 

I have seen no living plants of this genus and little herbarium material. Regel 
(1876) states that the frond-bases and cataphylls ("perulae") are at length deciduous, 
leaving a naked stem. By contrast, Hutchinson (1924) keys out the genus on its 
"trunk . . . covered with persistent leaf-bases and prophylls". This is a character of 
some importance in other genera. As stated above, the position of Microcycas is in 
some doubt. It does, however, show a number of resemblances to Zamia. 

8. Ceratozamia Brongn. Type species: C. mexicana Brongn. 

About four species in Mexico. Chromosome number: x = 8 (2m = 16). 
A distinctive genus. Its relationship is not certain, but appears to be rather 
distantly with Zamia. The fronds are more or less circinate in vernation. 

9. Zamia L. Type species: Z. floridana L. (lectotype, Hutchinson, 1924). 

About 30-40 species in the tropics of the New World. Chromosome number: x = 8, 
9 (2n = 16, 18). 

Like Cycas, Zaynia needs a modern monographic treatment, in this case by an 
American worker. Kegel's genus Aulacophyllum (Regel, 1876) is worthy of reconsidera- 
tion. Regel was a discerning worker and his treatment of cycad genera was sound. 
Aulacophyllum may be as good a genus as the equally neglected Lepidozamia. 

10. BowENiA Hook ex Hook. f. Type species: B. spectabilis Hook, ex Hook f. 
Two species in north-eastern Australia. Chromosome number: x - 9 (2n = 18). 
This is the only living cycad with truly decompound fronds, though Stangeria and 

Macrozamia may have forked pinnae. Discussed in detail in Part II (p. 109). 

Part II. The Zamiaceae of Australia. 

Historical. 
Early authors. 

Since Salisbury described Zamia spiralis (now Macrozamia spiralis (Salisb.) Miq.) 
in 1796, many authors have contributed to the systematics of the Australian Zamiaceae. 
No useful purpose would be served by a detailed account of their treatments and the 
progress, and at times regression, in the classification over the last century and a half. 
All the relevant references and an index of names may be found hereunder, in the 
formal systematic treatment. The more important landmarks and workers, however, 
may be briefly mentioned. 

Miquel established the genus Macrozamia in 1842, the three known species (under 
a single name) having previously been referred at first to Zamia and later to 
Encephalartos. This Dutch botanist studied the cycads in general for some thirty 
years and published numerous short and long papers on the group. Although many 
confusions are inevitably found in this early work, his studies are marked by a steady 
progress in understanding at both generic and specific levels, and his general discussions 
and contributions to cycad morphology are as praiseworthy as his taxonomic con- 
tributions. Unquestionably Miquel remains pre-eminent among taxonomic cycadologists 
and (the period taken into account) is overshadowed only by Chamberlain in the 
morphological field. Several of these studies (Miquel, 1861, 1863, 1868, 1869) are of 
particular importance. The first is a concise general review of the cycads of the world, 
while the second is a revision of the Australian members in which he reduced 
Macrozamia and the recently described Lepidozamia Regel to sections of Encephalartos 
and set up also the section Parazamia. The 1868 and 1869 papers are French and Dutch 
versions respectively of part of a series of morphological and taxonomic papers, of 
which the most important to us deals with the "Cycadeen van extratropisch Nieuw- 



72 THE FAMILIES OF CYCADS AND THE ZAMIACEAE OF AUSTRALIA, 

Holland", namely Macrozamia (restored here to generic rank, and embracing 
Lejndozamia as a section) and Bowenia. 

F. Mueller (latterly F. von Mueller) published numerous notes and brief com- 
mentaries between 1858 and 1889 (see references in formal section). On the credit side 
he made known a number of newly-discovered species, stimulated collection and 
increased the knowledge of distribution; unfortunately this was counterbalanced by 
serious confusion of quite dissimilar species (most notably M. miquelii with M. 
fawcettii) and thirty years of vacillation between recognition of Macrozamia and its 
inclusion in Enceplialartos. Each change of opinion was accompanied by a new com- 
bination or two. Mueller had little field knowledge of these plants, despite his residence 
in Australia. 

A notable cycadologist whose work is largely overlooked was Regel, who gathered 
a large collection of living plants at the then St. Petersburg Botanic Garden and paid 
particular attention to the most useful and significant vegetative characters of ilie 
genera. Among his many papers, two (Regel, 1857, 1876) are of special importance, 
dealing respectively with the new genus Lepidozamia (the distinctive characters of 
which subsequent authors have not properly appreciated) and with a general revision 
of all the Cycads. This little-known work is essentially more sound that later reviews 
of greater pretensions. 

J. D. Hooker (1863, 1872) described the distinctive genus Bowenia. A. De Candolle 
(1868) produced a useful monograph largely following Miquel. In the Flora 
Australiensis, BentlTam (1873), working in England from inadequate herbarium material, 
provided a better treatment than did Mueller on the spot, but, as he was himself aware, 
it is not very satisfactory. Successful cycad studies require years of familiarity with 
the plants. 

In a modest paper, Charles Moore (1884) gave the most realistic account yet 
published of the eastern species of Macrozamia (including Lepidozamia). Unlike other 
authors, Moore had the advantage of an extensive and critical field knowledge of most 
of the species. He also studied them side by side in the garden and laid the foundations 
of the present cycad collection in the Sydney Botanic Gardens. Apart from under- 
standable partial confusion of the species now defined as Macrozamia diplotnera, M. 
heteromera and M. sten-oviera (see p. 105) and neglecting minor matters of nomenclature 
with which he did not concern himself, the only flaws in his account arose from 
following Mueller with undue respect. Virtually the same arrangement was used in 
condensed form by Moore and Betche (1893). 

At various times, but especially in the Queensland Flora, F. M. Bailey (1902) 
reviewed the Queensland species. As with Moore, some field experience put Bailey In 
touch with reality, though not all of his species can now be maintained. 

From this encouraging position Maiden and Betche (1916) reverted to confusion 
by reducing all but one of the true Macrozamia. species of New South Wales to their 
inflated concept of M. spiralis, claiming that they were all connected by intergradation. 
This is not so. 

Schuster's monograph. 

As a greater anticlimax, Schuster (1932) produced his monograph in Das Pflanzen- 
reich. In his treatment of Macrozamia, despite its comprehensive scope and superficial 
aspect of detailed finality, Schuster so profusely introduces new and profound confusions 
in taxonomic concepts of every rank and in nomenclature, so blatantly contravenes the 
rules of priority, and so unreliably cites both synonymy and specimens that the work 
is quite egregious even for a compendium of such unequal quality as Das Pflanzenreich. 
Despite several years of unfortunately necessary detailed study of this work, I can 
form no idea of the principles, if any, which Schuster followed in arriving at his 
conclusions. Virtually identical plants, even the same specimens, are referred to 
entirely different species. Most diverse species are treated as subspecies, varieties or 
forms under a single (misapplied! ) name, often appearing more than once in quite 



BY L. A. S. JOHNSON. 73 

different branches of this elaborate false hierarchy, while species closely allied to some 
of these are given full specific rank. For instance, Schuster places materials of M. 
heteromera as (i) (the typical variety of) an independent species, but there confused 
with two other species; (ii) a form of a variety (the tyj)e form of which is a different 
species) of this same species; (iii) a form of a variety of a subspecies of "M. tridentata" , 
this form being a mixed concept of two species, the variety a mixed concept of four 
other species, and so on through the higher categories. With one or two partial 
exceptions, his arrangement and circumscription of all taxa from sections down to 
forms bear almost no relation to the real affinities and the distinctions between the 
species in nature. The result is a jumble which almost defies disentanglement. It is 
hard to see why Schuster did not follow the earlier treatments which, though imperfect, 
should have been a useful guide. 

In most branches of science worthless work may be forgotten and need not hinder 
progress; taxonomy is bedevilled by the requirement that no validly published name 
may be ignored. Thus taxonomists spend much of their time in unprofitable 
antiquarianism rather than scientific study. So that future students of Australian 
Zamiaceae need not repeat much of this labour, and perhaps to help those in charge 
of overseas cycad collections who have naturally had recourse to Schuster's monograph 
in naming their plants, I have provided a table (Table 1) interpreting his usage. 

Further details may be found in the synonymy of the various species in the formal 
section, below. In using this tabulation one should bear in mind that in any subdivided 
taxon Schuster's method was to describe and list first all material which he treated as 
typical and to follow this by his first named subsidiary category; (supposedly) typical 
subdivisions were not named in the lower rank. For example, his "Macrozamia 
ti'identata" has two named and numbered subspecies, but these are preceded by a 
description and citations applying only to what one must consider as the "typical" sub- 
species (though the term has little meaning here; how Schuster regarded it one can 
only guess). Similarly his first subspecies includes three named varieties exclusive of 
the "typical" variety and so on. In Table 1 the, total ambit of his inclusions ("concept" 
is hardly the word) under any one taxon must be obtained from the sum of the 
interpretations of its subdivisions. 

In order to achieve finality in this tedious matter the identity of these "synonyms" 
and specimens has been checked with great care, and the reductions, as amplified in 
the formal section, may be accepted as authoritative according to the concepts of the 
present treatment. All cases of residual doubt or ambiguity are clearly indicated. 
Column 2 of the table gives the present correct position of the type(s) of the 
basionyms (and absolute nomenclatural synonyms) of Schuster's names, not the actual 
basionym itself. The latter may be determined from the synonymy in the formal 
section. Similarly, column 4 lists the correct positions of the types of the various 
alleged synonyms cited by Schuster, excluding the absolute synonyms. 

Schuster's descriptions are frequently largely copied or translated from those of 
other authors and do not necessarily apply to the specimens or synonyms cited with 
them. 

Present Study. 
Scope. 

This paper aims to present a fully-documented and definitive taxonomic revision ol, 
and commentary on, all species of the three Australian genera of Zamiaceae. Amongst 
these the New South Wales species have received most attention, and indeed have 
most needed it. 

No new cytological or anatomical investigations have been made, but the results 
of such studies are taken into account in forming taxonomic judgments. General 
ecological observations have been made and their taxonomic significance assessed. 
Economic and horticultural aspects lie outside the scope of this paper, but have been 
borne in mind in the course of the work. Evolutionary and theoretical questions are 
discussed when relevant. 



74 



the families of cycads and the zamiaceae of australia, 
Table 1. 





Correct Name in the Present Revision. 

[Macrozamia spp. belonging to sect. Parazamia marked (P), 

others belong to sect. Macrozamia.] 


Name Used by Schuster (1932) 
(spellings corrected). 


Basionym 

(present correct 

position). 


Specimens Cited. 


" Synonyms " Cited 

(excl. Invalid 

Names and 

Absolute Synonyms) 

(present correct 

position). 


M. triderdata 

(Used in contravention of priority, since 
Schuster's circumscription included 
the types of several earlier names, viz. 
Z. spiralis Salisb., Z. pallida Salisb., 
Z. pungens Ait.) 


Macrozamia or En- 
cephalartos sp., un- 
identifiable (juv.) 
( = Z.tridsntataWmd.) 


M. miquelii. 
M. communis. 

(and perhaps M. 
Iwida (P) by in- 
ference). 


M. miquelii. 

M. communis. 

M. spiralis (P). 

Zamia, 

EncepJialartos 

or Macrozamia sp. ? 

(Z. pallida Salisb.). 


ssp. mountperriensis (" mountperryensis ") 


M. miquelii. 


M. miquelii. 




var. miquelii (nom. Ulegit.) 


M. miquelii (lectotype) 
and M. fawcettii (P). 


M. miquelii. 


M. miquelii. 
M. fawcettii (P). 


f. milkaui (nom. illegit.) . . 




M. miquelii (juv.). 
M. oxEnceph. sp. (juv.) 
(=Z.tride7datay>l\aA.) 


M. or Enceph. sp. 
{ = Z. tridentata 
Willd.). 


f. oblongifolia 


M. miquelii. 


M. communis (juv.). 




var. mackenzii 


M. miquelii. 


M. miquelii. 




var. douylasii 


M. miquelii. 


M. miquelii. 




ssp. cylindrica 


M. miquelii. 


M. miquelii. 
M. communis. 
M. moorei. 




var. secunda 


M. secunda (P). 


M. secunda (P). 
M. pauli-guilielmi ssp. 
plurinervia (P). 




var. corallipes 


M. spiralis (P). 


M. spiralis (P). 


M. spiralis (P). 


f. wallsendensis 




M. communis (semi- 
juv.). 




f. dielsii 




M. spiralis (P). 




f. vavilovii 




M. communis ? 
or M. miquelii ? 




var. pungens 


Encephalartos sp. ? 
(obscure, not Aus- 
tralian). 


M. communis. 
M. miquelii. 
M. moorei 
M. lueida (P). 


Zamia. 
Enceph. or 
Macroz. sp. ? 
(Z. pallida Salisb.). 


f. hillii 




M. miquelii 
or M. communis ? 
(or even M. riedXei ?). 


M. miquelii 

or M. communis ? 
(or even M. riedlei ?) 


f. diplomera 


M. diplomera (lecto- 
type) 
and M. heteromera (P). 


M. diplomera. 

M. heteromera (P). 





BY L. A. S. JOHNSON. 
Table 1. — Continued. 



75 





Correct Name in the Present Revision. 

[Macrozumia spp. belonging to sect. Parazamia marked (P), 

others belong to sect. Macrozamia.] 


Name Used by Schuster (1932) 
(spellings corrected). 


Baslonym 

(present correct 

position). 


Specimens Cited. 


" Synonyms " Cited 

(excl. Invalid 

Names and 

Absolute Synonyms) 

(present correct 

position). 


M. moorei 


M. moorei. 


M. moorei. 




M. flexuosa 


M. pauli-guilielmi ssp. 
flexuosa (P). 


M. pauli-guilielmi ssp. 

flexuosa (P). 
M. p.-g. ssp. pluri- 

nervia (P). 




M. heteromera 


M. heteromera (P) 
(lectotype) 
and M. diplomera. 


M. heteromera (P). 
M. stenomera (P). 
M. diplomera. 




var. dieranophylloides 




M. stenomera (P). 




var. tenuifolia (nom. illegit., later 
homonym). 




M. stenomera (P). 




f. harmsii 




M. heteromera (P). 




M. pauli-guilielmi (" pauli-guilelmi ") 


M. pauli-guilielmi 
ssp. pauli-guilielmi 
(P). 


M. pauli-guilielmi 

ssp. pauli-guilielmi 

(P). 
M. p.-g. ssp. flexuosa 

(P.) 
M. p.-g. ssp. pluri- 

nervia (P). 


M. pauli-guilielmi 
ssp. pauli-guilielmi 
(P). 


M. fawcettii 


M. fawcettii (P). 


M. fawcettii (P). 

M. pauli-guilielmi 

ssp. flexuosa (P). 
M. p.-g. ssp. pluri- 

nervia (P). 




M. platyracMs (" platyrhachis ") . . 


M. platyrachis (P). 


M. platyrachis (P). 




M. denisonii . . 

(Used in contravention of priority.) 


Lepidozamia 

peroffskyana. 


L. peroffskyana. 
L. hopei. 


L. peroffskyana. 


var. hopei 


L. hopei. 


L. hopei. 




M. preissii 

(Used in contravention of priority.) 


M. riedlei. 


M. riedlei. 

M. m%cdonnellii. 


M. riedlei. 

M. macdonnellii. 


ssp. dyeri 


M. riedlei. 


M. riedlei. 




M. Sect. Polyorientales (nom. illegit.) 


As to type sp. 
M. ^Macrozamia. 


M. ^Macrozamia (part). 
M. ^Parazamia (part). 




Subsect. Atlenuatae (nom. illegit.) 


M. ^Macrozamia. 


M. §Mocrozamia (part). 
M. ^Parazamia (part). 




Subsect. Acutae 


M. ^Parazamia. 


M. ^Parazamia (part). 




Subsect. Curvatae 


M. ^Parazamia. 


M. ^Parazamia (part). 




M. Sect. Monoorientales (nom. illegit.) . . 


Lepidozamia (Genus). 


Lepidozamia (Genus). 




M. Sect. Monooccidentales (nom. illegit.) 


M. ^Macrozamia. 


M. ^Macrozamia (part). 





76 



THE FAMILIES OF CYCADS AND THE ZAMIACEAE OF AUSTRALIA, 



Table 1. — Continued. 





Correct Name in the Present Revision. 

[Macrozamia spp. belonging to sect. Paramimia marked (P), 

others belong to sect. Macrozamia.] 


Name Used by Schuster (1932) 
(spellings corrected). 


Baslonym 

(present correct 

position). 


Specimens Cited. 


" Synonyms " Cited 

(excl. Invalid 

Names and 

Absolute Synonyms) 

(present correct 

position). 


Boivenia spectabilis . . '. 


B. spectabilis. 


B. spectabilis. 




var. serrulata 


B. .serrulata. 


B. serrulata. 

B. spectabilis {?). 


B. serrulata. 



Basis. 

The materials used are: (1) The herbarium and museum collections of the New 
South Wales National Herbarium (NSW), together with certain material from the- 
Botanic Museum and Herbarium, Brisbane (BRI), the National Herbarium, Melboui'ne 
(MEL), the Museum of Applied Arts and Sciences, Sydney (TECH), and the Botany- 
Department, University of Sydney (SYD). Since the outstanding problems were 
centred in New Soutli Wales, only selected specimens from the other State herbaria 
were examined. Their full collections will, of course, fill in details of distribution, but 
are not likely to affect the conclusions, at least without concurrent field study. (2) The 
living cycad collections in the Royal Botanic Gardens, Sydney. (3) Natural populations 
of the following species: Lepidozamia peroffskyana, Macrozamia moorei, M^ 
macdonnellii, M. commimis, M. diplomera (limited observation), M. lucida. M. spiralis, 
M. secunda, M. heteromera, M. stenomera and M. 2}(i,uU-guilieliui (especially ssp. 
plurinervia) . The only species not examined either in the field or the garden are 
L. hopei. M. platyrachis and M. riedlei. The last certainly needs some field study. 
(4) Published and unpublished illustrations and descriptions. Information solely 
derived from such descriptions, except when they are very clearly fully reliable, is 
specified as such in the text. 

Outlook. 

Since systematists are notoriously individualistic, it is always helpful to appreciate 
a particular worker's standpoint and philosophical approach to the subject. Underlying 
the present treatment is the conviction that taxonomic classification, in one way or 
another, should reflect, if not express, the results of evolution. When little palaeonto- 
logical evidence is available one can do this only by inference, using the accumulated, 
knowledge of what is known to have happened in other and especially related groups 
of organisms, and evidence of every possible kind must be evaluated and taken into 
account. Bearing in mind genetic, evolutionary and ontogenetic principles, one must 
try to determine the most probable relative phylogenetic relationships and divergences 
of the various individuals, populations and groups of populations, and decide their 
taxonomic status accordingly. The success of this method depends greatly on the 
student's knowledge and insight, but on a balance of probabilities it leads to mor& 
satisfactory conclusions than those obtained by the static or formalistic approach,, 
which rejects evolutionary speculation and depends solely on morphological differences 
and resemblances in their own right. It is well known that similar conclusions may 
often be reached from these really quite fundamentally different starting points. That 
they quite often lead to different conclusions is frequently forgotten. 

I have, of course, used the customary, chiefly macroscopic and external, morpho- 
logical characters in the keys and descriptions. However, characters as such were not 
regarded as the basis of taxonomic decisions, but rather as indices of the natural 
populations or groups of populations between which, on grounds of evolutionary 



BY L. A. S. JOHNSON. 77 

probability, if not direct evidence, we can infer relationship. It is often held that the 
best taxonomic characters are those which are selectively neutral. The existence of 
truly selectively neutral characters is dubious; practically every characteristic of a 
line of organisms is in some sense adaptive, within the range of genetic possibilities 
open to the particular group. Though selection may operate chiefly on some genetically 
and developmentally correlated characters rather than those we observe as taxonomic 
differences, any deleterious effects will tend to be minimized by adaptive compensation 
(Stebbins, 1950, pp. 121-123). There are, of course, many genetic and physiological 
balances and counterbalances, multiple genetic effects, and time-lags, and some 
characters do appear fixed because they are adapted not to the external environment, 
but rather to the complex of balanced organization within the organism itself in such 
a way that any uncoordinated change is disadvantageous. 

Such characteristics are indeed of great taxonomic value and phylogenetic sig- 
nificance, but they are not confined to any particular set of organs, and one cannot 
uphold the often expressed view that, in the vascular plants, leaf and stem characters 
have less taxonomic value than reproductive characters. Variations in over-all shape 
and size, whether due to genie or direct environmental differences, cause little upset 
in the complex processes of the plant; such variations are indeed most common in 
vegetative organs, but may be found in the reproductive system as well. For instance, 
cone and sporophyll size varies considerably, both according to conditions and 
(apparently) genetically, within single species of Macrozamia. 

In short, my attitude to taxonomic theory in relation to evolution is in general 
similar to that of Simpson (1945, 1953), with the difference that my emphasis is on 
the classification of the present-day plants in the light, where possible, of the evolution 
of their ancestors. To fit arrays of organisms widely separated in geologic time into a 
single classificatory system is often impossible. Organisms are related both to their 
contemporaries and to their forebears, but the relationships are not comparable. 

Concepts. ; 

The genera of cycads are too clearly distinct and too coherent in themselves to 
cause much argument as to status. This applies as well to Lepidozamia as to the 
currently-accepted genera. Botanists (except Regel) have simply neglected to look at 
it or think critically about it. The characters by which genera differ may be anything 
at all, according to the particular case. 

A species, in this normal sexual group, is taken to be a population or group of 
similar populations effectively isolated in nature hy any means which prevents a degree 
of interbreeding sufficient to cause large-scale merging in characters with related but 
genotypically and phenotypically consistently different populations. This is a crude 
-but practical definition. The category cannot be precisely defined, and populations 
may occur which could reasonably be called either species or subspecies. 

A subspecies I take to be a geographically or ecogeographically characterized 
population (or group of populations) consistently and recognizably different in several 
respects from, but intergrading to a considerable degree with neighbouring populations. 
A subspecies must be coherent as a population, that is, there must be much more 
breeding within the subspecies than between them. A variation of the definition is 
that if subspecies A and B merge, but a third very similar group C is isolated, C may 
be regarded as another subspecies rather than a distinct species. 

The category of variety is not used for reasons given below (p. 108). 

Morphological terminology. 

Much confusion and false argument could be saved by a precise rationalized 
terminology for comparative morphology, to be clearly distinguished where necessary 
from non-committal, purely descriptive terms. In the present case one must bear in 
mind that most structures of cycads are not homologous with those bearing similar 
names in the Covda,ite-Ginkgo-Ej)hedra (Bames. 1952)-Taxad-Conifer (Florin, 1948, 1954) 



/O THE i'AMILIES OF CYCADS AND THE ZAMIACEAE OF AUSTRALIA, 

group* with which they are commonly grouped as "Gymnosperms". (It is high time 
that this term and "Pteridophyta" were relegated to the history of botany. Their 
continued use misleads not only students but learned botanists into the belief that they 
apply to natural groups.) The cycads, of course, belong to the Pteropsida, in company 
with the true fern groups both ancient and modern, Pteridosperms, the Glossopteris 
(Scutum) group (Plumstead, 1956), Bennettitales, Caytoniales, Welwitschia, Gnetuvi- 
(Eames, 1952, p. 96) and, with little doubt, the Angiosperms. It does not follow that 
every organ called by the same name in these groups is homologous, though many are 
so, at least in a broad sense. 

In general I have used terms which either (a) are well known to be descriptive 
and without comparative morphological significance, or (b) have a definite morpho- 
logical sense which applies correctly to the cycads. However, a few terms fall between 
these categories and need comment. 

Frond is used rather than "leaf", to indicate that these organs are generally com- 
parable to the fronds of a fern, that is to say, they clearly show their derivation from 
the branch-systems of early Pteropsida, but have become two-dimensional and of quite 
determinate and predominantly intercalary growth. The leaves of Angiosperms may be 
homologous, but are certainly rather dissimilar structures. The "leaves" of non- 
pteropsids are either not at all or only in a very broad sense homologous. 

Petiole is used as a descriptive term, without morphological significance, for the 
frond-stalk proximal to the first pinnae. Stipe (s), used for the equivalent of the petiole 
in ferns, is traditionally used for the stalk of the megasporophyll in the cycads. 

RhacJiis is used for the frond-axis from the lowest pinnae onwards. 

Pinna is used for the primary frond-segments. Pinnule is used only for the 
ultimate segments of Bowenia, whether these are borne on a secondary rhachis or on 
the end of the primary rhachis. The term is not applied to the segments of the 
bifurcating pinnae of some species of Macrozamia. "Leaflet", like "leaf", is not used. 

Cataphyll, rather than "scale-leaf", which is scarcely apt here, is used for "reduced" 
leaves. 

Cone (in Latin descriptions conus rather than strohilus) is used in a non-committal 
descriptive sense. The cycad cone is not at all equivalent to the female cone of conifers 
(a compound strobilus), nor is it really equivalent to the male cone of the same group, 
since the sporophylls are not homologous. Both terms "cone" and "strobilus" are used 
for a number of non-homologous structures; for comparative purposes they need 
replacement by special terms of obvious significance. I suggest strohoid for the cycad 
"cone", strobe for the male (or single unit of a female) conifer cone, synstrobe for the 
female conifer cone, anthoid for the Bennettitalean bisexual structure, but I shall 
pursue this matter elsewhere. Further simple, euphonious terms of obvious meaning 
are necessary for "cones" of Lycopods, Lepidodendrids, and the Sphenopsida. Such a 
terminology would add to precision, while retaining a degree of simplicity of language. 

Cone-stalk-f (axis coniger) is preferable to the usual "peduncle" (which it is not, 
since it bears the approximate equivalent of a single flower) or "pedicel" which suggests 
a more slender structure. 

Sporophyll is used in preference to "cone-scale". The structure is quite different from 
the complex female cone-scale of the conifers. Admittedly it is not really homologous 
with the sporophyll of the male conifer strobilus and certainly not with the "sporophyll" 
of the Lycopods. Again the term is used in the non-committal sense of a simple (or 

* For this great branch of the vascular plants the terms Pityopsida or Pityophyta 
(according- to rank) are suggested as etymologically preferable to Coniferopsida and 
Coniferophyta (cf. Arnold, 1948), which are Latin-Greek mongrel terms. Pitys is Greek for 
a pine and' is also the name of a genus of Palaeozoic near-conifers. Since the I. C.B.N, 
fortunately does not enjoin respect of priority in naming higher groups, we may surely choose 
to use euphonious, well-formed and meaningful names. 

t Peduncle Is used in my forthcoming account in the Flora of New South Wales, due to 
its use in the earlier ms. of that work. 



BY L. A. S. JOHNSON. 79 

apparently so) sporangium-bearing organ more or less equivalent to a "leaf" of the same 
plant. 

Melville (1957) has introduced an elaborate special terminology to describe the 
various angles and faces of the sporophylls of Enceplialartos. These terms certainly 
facilitate description of the very geometrical sporophyll-ends in that genus, where, 
moreover, these details are of taxonomic importance. In the Australian genera, however, 
they are not necessary and frequently not applicable. I have simply referred to the 
sporophyll-end, its wing and its spine. The last two (in Macrozamia) are actually 
continuations of the margins and the end of the lamina of the sporophyll. The 
sporophyll-end is the swollen portion distal to the stipe, and the descriptions apply 
especially to its outer surface on the cone, which really includes both adaxial and 
abaxial faces. 

Male and female (rather than "microsporangiate", etc.) are applied freely to the 
cones and sporophylls (and plants). The sexual differentiation, which is finally expressed 
in the gametes, is carried back into the sporophyte, and I see no more objection to 
the use of sexual terms for a plant sporophyte than for an animal diplont, such as a 
man. Megasporophyll is, however, used at times (not in the descriptions). 

Ovule is used for the megasporangium with its integument. The term is a descrip- 
tive one not implying complete homology in different groups. The same applies to seed. 

All such terms as "flower", "stamen", "staminate", and "anther", though they have 
been widely used, especially on the continent of Europe, in cycad and conifer descrip- 
tions, are better confined to the angiosperms. 

Taxonomic criteria. 

Certain characters have proved useful for the recognition and formal definition of 
the taxa, though these are not necessarily more essential than others less useful. The 
keys and descriptions herein merely differentiate, circumscribe and partially describe 
the natural populations in terms of the readily observed external morphology. No 
character, as such, is essential to any one taxon. The correct identification of an 
individual is the determination of the natural population of which it is part, not the 
key-characters which it possesses. The latter are merely a generalization, as accurate 
as possible. For example, plants of Macrozamia com-munis may rarely be found with 
spirally twisted fronds, contrary to the key and descriptions. These aberrant plants 
are still constituents of the population defined as M. communis and the key (as a 
99-f- per cent, generalization) is not wrong. The dimensions given, in particular, are 
not to be considered absolute; cycads are especially variable in this respect. Never- 
theless, the dimensions used in the keys apply to all the specimens actually examined, 
and there should not be many departures from them. 

The characters which have been found useful will be evident from the keys and 
the discussions under the genera and sections. They need not be detailed here. In 
all the Australian Zamiaceae, many vegetative characters are quite as consistent as 
those of the cones, and these are given prominence in the keys. They will be especially 
useful in Macrozamia section Parazamia, in which cones are irregularly produced and 
frequently only vegetative individuals may be found; the cones are much the same 
throughout this group. 

Juvenile and semi-juvenile plants are usually indeterminable unless their geographic 
origin is known. Only adult plants (old enough to bear cones) are described herein 
(see below, p. 89). The shapes of rhachis and petiole change somewhat on drying, as 
the parenchymatous tissue shrinks and the sclerenchyma is externally revealed as 
ridges. As far as possible both living and dried material is described. 

Morphological and anatomical studies. 

The detailed morphology, anatomy and life-history of the cycads have attracted 
a great deal of attention, especially from Chamberlain and his students. This work 
of great importance in the comparative morphology of the vascular plants generally. 



80 THE I'AMILIES OF CYCADS AND THE ZAMIAGEAE OF AUSTRALIA, 

cannot be discussed here, and it is assumed that any serious student of cycads is 
familiar with its chief conclusions. The most notable Australian study is that of 
Brough and Taylor (1940) on Maerozamia communis ("M. spiralis" in their usage). The 
bibliography of their paper will provide an entry to the morphological literature. 

Glu-oniosome numbers. 

All the Australian species cytologically examined have the chromosome number 
2n = 18, except for an early count (n = 12) on M. riecllei (q.v.), which needs confirma- 
tion. The numbers are cited after the descriptions of the genera and species. No 
chromosome count has yet been made in Lepidozamia. 

Distribution. 

The detailed distribution is indicated under each species and the present distribu- 
tion patterns, which show much vicariism, are discussed under the genera.* Not a 
great deal can be said about past distribution (but see the discussion under Lepidozmiiia 
hopei, p. 86). Maerozamia section Maerozamia must have extended across the southern 
half of the continent during the Tertiary, though not necessarily continuously at any 
one time. Sectioii Parazamia and the other two genera are confined to eastern Australia. 
The nearest relationships of Lepidozamia and Maerozamia appear to be with the African 
Encephalartos, but since the cycads in general are a relict group with scattered 
surviving genera, not too much weight can be placed on this. Bowenia is very 
different from the other Australian Zamiaceae, and is probably closer to some of the 
American genera. Again, the phytogeographic significance of this is hard to evaluate 
in an ancient and fragmented group. 

Table 2 gives the distribution by States. No Zamiaceae occur in Victoria, Tasmania 
or South Australia. It must be kept in mind that this table excludes Cyeas (Cycadaceae 
s. str.) which has species in the tropical parts of Western Australia, the Northern 
Territory and Queensland. Distribution maps are given under the respective genera 
and sections, below. 

Speciatio7i. 

While little can be said about details of the evolution of the genera and sections, 
the distribution patterns suggest that the more recent speciation in the Australian 
Zamiaceae has been dependent on geographic isolation of segments of formerly con- 
tinuous populations, associated with differentiation in response to environmental 
changes. This classical ecogeographic speciation is perhaps not quite complete in some 
cases especially in Maerozamia section Parazamia (see p. 100). 

There is no evidence of recent polyploidy in any of the living cycads. The universal 
dioecism would make successful polyploidy unlikely, though not impossible. Outbreeding 
Is, of course, complete. Some local ecotypic differentiation is found' in the more wide- 
spread species (see discussions under the individual sections and species of Maerozamia) . 

Hybridism. 

As would be expected in anemophilous, dioecious plants, all (so far as known) with 
the same number of chromosomes, some natural hybridism is found in Maerozamia 
where the species come into contact. Some introgressive hybridization is possible in 
the case of M. diplomera (q.v.), but in general the hybrids do not appear to be of much 
importance. A few spontaneous hybrids have arisen in the Maerozamia collections in 
the Sydney Botanic Gardens. 

Particular cases are discussed under the respective parental species (below, or in 
Johnson, in press), but the following apparent hybrids have been found in the field or 
in herbarium collections: M. diplomera x M. lieteromera, M. lucida x M. moorei, M. 
heteromera x M. seeunda, M. lieteromera x M. stenomera, M. pauli-guilielmi ssp. 



* The subdivisions of New South Wales are defined as in the Flora of New South Wales 
(Andetson, in press). 



BY L. A. S. JOHNSON. 



81 



plurinervia x M. stenomera (or perhaps M. heteromera) . Others may be expected. 
All of these have been recognized by obviously intermediate frond characters 
and their occurrence as single or infrequent individuals in mixed communities of the 
apparent parents. Cones have not been available and no progenies could therefore be 
raised. 

Ecology. 

The particular ecology is discussed under each taxon. Most species grow in 
Eucalyptus (sclerophyll) forest, but Lepidozamia hopei (and to some extent L. 
peroffskyana) and Bowenia spectaMlis may be found in certain sites within or near 

Table 2. 





Qld. 


N.S.W. 


N.T. 


W.A. 


Number of genera 


3 


2 


1 


1 


Number of species 


9 (2 : 5 : 2) 


12 (1 : 11 : 0) 


1 


1 


Zepidozamia 










hopei 


+ 








peroffskymm 


+ 


+ 






Maerozamia 










(Sect. Maerozamia) 










moorei 




+ 


+ 






riedlei 










+ 


macdonnellii ' . . 








+ 




miQuelii 




+ 


+ 






communis 






+ 






diplomera . . 






+ 






(Sect. Parazamia) 










platyraekis 




+ 








lucida 




+ 


+ 






spiralis 






+ 






secunda 






+ 






heteromera 






+ 






stenomera 






+ 






pauli-guilielmi 




+ 


+ 






ssp. p.-g. 




(+) 


( + )* 






ssp. plurinervia 




(+)• 


(+) 






ssp. ffexuosa . . 






(+) 






fawcettii 






+ 






Bowenia 










speciahilis . . 


+ 








serrulata 




+ 









* Intergrades only. 

rainforest. Only one species, M. macdonnellii, occurs under really arid conditions, and 
even this is confined to the refugium of the Central Australian range system where 
conditions are somewhat better than elsewhere in the interior. The species of 
Macroza^nia, in contrast to the other two genera, are usually associated with other 
sclerophyllous species of the "Australian" floristic element. The plants are rather slow- 
growing, but by no means as excessively so as sometimes stated (see Lepidozamia 
peroffskyana, p. 86). Apogeotropic roots regularly form coralloid end-systems containing 
Cyanophyta (Nostocaceae: Anaiaena and perhaps Nostoc) (Fritsch, 1945, pp. 872-4). 
These blue-green algae may be in a symbiotic relationship with the cycad plant. 
Nitrogen-fixing bacteria (e.g. Azotobacter) also occur in these nodules in a number of 
cycads (refs. in Thieret, 1958). 

Economics. 

Like other cycads, the Australian Zamiaceae contain toxic substances (Hurst, 1942; 
Webb, 1948; Gardner and Bennetts, 1956) which are at times responsible for poisoning 



82 THE FAMILIES OF CYCADS AND THE ZAMIACEAE OF AUSTRALIA, 

of stock, especially cattle. For this reason they have been killed out in certain areas. 
In view of their great botanical interest and the low pastoral quality of much of the 
land on which they occur, their destruction is particularly unfortunate. Nevertheless, 
in the past some botanists, especially in Queensland and Western Australia, have joined 
the agriculturists in issuing publications recommending methods of poisoning the 
plants. 

The starchy endosperm of the seeds, and sometimes the pith of the caudex, served 
the aborigines as food, as Gycas still does in Arnhem Land (Specht, 1958), after washing 
and roasting to remove the poison. Various cycads are used similarly in other con- 
tinents. Thieret (1958) gives a comprehensive review of the literature on the economics 
of cycads, but much of the cited information is out of date. 

In horticulture, various species are to be seen in home and occasionally public 
gardens in Australia, but many species have been cultivated under glass in European 
collections, especially during the nineteenth century. The description of such plants 
(usually from vegetative and often semi-juvenile material) has been responsible for 
much of the extraordinary confusion of the nomenclature. 

Taxonomic Revision. 
Note. 

This paper will be followed very shortly by the publication of the first parts of the 
Flora of Neio South Wales (ed. Anderson, in press). My treatment of the Zamiaceae 
forms Part 1 of this work. For reasons of space, descriptions (except Latin descriptions 
of new taxa), detailed synonymy and lists of specimens of the New South Wales species 
are not given in the present paper, but will be found in the Flora. Reference to the 
latter is made where necessary in the text hereunder. A complete index of synonyms, 
however, is given below (p. 113). 

For most purposes the keys herein will be sufficient for identification, especially if 
the geographical origin of the material is known, but for critical investigation this 
revision and the Flora must be used in conjunction. I regret this necessity. 

Family ZAMIACEAE Reichenb. 
Sporophylls of both sexes in definite simple cones, of determinate growth. Cones 
dioecious, rather large, axillary or terminal, with numerous scale-like sporophylls 
arranged spirally or apparently in vertical rows; male sporophylls (microsporophylls) 
with many small, globose sporangia more or less clustered in groups (sori) in two 
collateral sometimes confluent areas on the undersurface, sporangia dehiscent by slits; 
female sporophylls (megasporophylls, macrosporophylls) simple, consisting of a barren 
stipes and an expanded and thickened end, the latter bearing two (rarely three in 
Lepidozaviia) sessile ovules on its axis-facing margins ("inner surface"). Pollen wind- 
borne; fertilization by naked multiflagellate spermatozoids. Seeds large, from sub- 
globular to subcylindrical or variously angled, with a broad chalazal area and more or 
less apiculate micropylar end, seed-coat with a fleshy outer layer and a woody inner 
layer; (haploid) endosperm present, cotyledons two, germination hypogeal. 

Somewhat palm-like plants; stem subterranean to tall, not or little branched, with 
a crown of spirally arranged pinnate or bipinnate leaves (fronds) and interspersed 
rudimentary leaves (cataphylls), leaf-bases persistent and clothing the stem, or 
deciduous. Fronds usually straight and folded in bud (circinate in Ceratozamia of 
Mexico); leaf-segments without a midrib, with few to many longitudinal nerves, 
straight, not circinate, in bud. 

A family of eight genera and about 80 species, with a scattered distribution in 
tropical to warm-temperate parts of N. and S. America, Australia and Africa. 

For a key to the tribes and genera see Part I, p. 66. 

Australian members: Tribe Encephalarteae: I. Lepidozamia; II. Macrozamia. Tribe 
Zamieae: III. Boioenia. 



BY L. A. S. JOHNSON. 8S 

I. LEPIDOZAMIA Kegel. 

Kegel in Bull. 8oc. Imp. Nat. Mosc. (1857), n. 1, 182, and in Gartenfl. 6 (1857*), 11, 
ibid. (1870), 227, (1875), 42 and (1876), 4, also in Acta Hort. Petrop., 4 (1876), 294; 
Miq., Prodr. Syst. Cycad. (1861), 10; A.DC, Prodr., 16, pt. 2 (1868), 547 (as "Genus, 
dubium"). 

Typification: L. peroffskyana Kegel. Type Species (the sole original species). 

Synonymy : 

CataKidozamia W. Hill in Gard. Chron. (Nov., 1865), 1107 (type sp.: C. hopei W. 
Hill). [This is to be regarded as a combined "generico-specific" description, since only 
one valid species was included, though "C. macleayi" was mentioned as a nomen nudum. 
Schuster (1932, p. 86), who seems to have cited many publications without seeing them, 
has followed Index Keioensis in giving the author as T. Hill. The generic name evidently 
circulated amongst European gardeners (e.g. Haage and Schmidt) as "Katakidozamia'\ 
and is thus cited, in synonymy, by Kegel (1876, both papers), but without reference to 
Hill. Schuster mis-cites it as "Katikidozamia" .1 

Encephalartos Lehm. sect. Leindozamia (Kegel) Miq. in Yersl. Meded. Koninkl. 
Akad,. 15 (1863), 371. 

Macrozamia Miq. sect. Lepidozamia (Kegel) Miq., in Arch. Neerl., 3 (1868), 253, 
and in Versl. Meded Koninkl. Akad., ser. ii, 3 (1869), 56. 

Macrozamia sect. Monoorientales Schuster, Pflanzenr., IV, i (1932), 88, nom. illegit. 
(exactly equivalent to the above). 

Palm-like plants with a usually unbranched trunk clothed by the persistent leaf- 
bases, often tall, all parts shortly pubescent to tomentose when young, but some 
glabrescent with age. Fronds numerous, large, simply pinnate, not twisted. Cataphylls 
abundant, fleshy, linear-subulate, shortly tomentose, mostly in series alternating with 
the successive crowns of fronds. Base of petiole swollen, shortly tomentose. Pinnae 
numerous, spreading, inserted more or less alternately along the adaxial midline of the 
rhachis, falcate, slightly narrowed but not callous at the base, stomata confined to lower 
surface. Cones of both sexes almost or quite sessile, large, axillary among the cataphylls 
of the crown, the spirally-arranged sporophylls tomentose at the ends, not spiniferous. 
Male sporophylls with a linear-spathulate fertile region and a more or less triangular- 
rhomboid, laterally expanded end, deflexed in spiral series at sporangial dehiscence. 
Female sporophylls with an elongated, terete to subangular stipes and a dorsiventrally 
biconvex, laterally expanded, somewhat deflexed, shortly tomentose end bearing two 
or sometimes three ovules on the inward-facing margins (inner surface), tapered into 
the flattened, more or less acute, at first deflexed, but on drying more or less upturned, 
ultimate tips. Outer seed-coat fleshy, red or yellowish, inner coat hard. 

Endemic in the eastern coast region of Australia, with two species, one in tropical 
Queensland, the other in subtropical Queensland and the north coast of New South 
Wales, in wet sclerophyll forest and around rainforest (Text-figure 1). 

Except by Kegel himself, Lepidozamia has been generally placed under Macrozamia 
ever since its Australian origin was known (the type species was described from a 
garden plant in the then St. Petersburg, of unknown origin, and was at one time thought 
to be from Mexico). However, it is quite as distinct as any other genus of Zamiaceae, 
and I can see no especially close affinity with Macrozamia. 

Encephalartos, Macrozamia and Lepidozamia would all appear about equally 
similar — and dissimilar — and if they appear to stand together apart from other 
Zamiaceae, this is perhaps due to the absence of certain advanced or specialized 
characteristics, rather than to any very positive features in common. They can, of 
course, be made to stand together in a key, but keys are an unsafe guide to relationships. 

Lepidozamia is readily distinguished from Macrozamia by the almost or quite 
sessile cones, the quite differently-shaped sporophyll-ends which lack the characteristic 

* Later than the first-cited reference. I have not seen the four Gartenflora references, 
which are quoted from Kegel's second 1876 paper. 



84 THE FAMILIES OF CTCADS AND THE ZAMIACEAE OF AUSTRALIA, » 

spinescent modiflcation of Macrozamia and moreover are closely tomentose-pubesceni 
with short hairs, the short swollen rather than laterally expanded and shortly rather than 
silky-woolly tomentose petiole-bases, the fleshy subulate shortly-tomentose cataphylls, 
the structurally more involute rhachis with median insertion of the falcate pinnae 
which lack callous bases, the straight torque-free vernation of the fronds and the more 
open crown with frond-series markedly interrupted by wide cataphyll-series. Further- 
more, as Thomas and Bancroft (1913) and, in a more exhaustive study, Cookson (1953) 
have found, the epidermal anatomy of the species referred to Lepidozamia differs 
strikingly from that of Macrozamia proper. In Lepidozamia the long axes of the 
epidermal cells are oblique or transverse to the long axis of the pinna, in contrast to 
the longitudinal orientation in Macrozamia. 



Text-flg. 1. — Distribution of Lepidozamia: 1: L. hopei; la: site of L. hopeites fossils; 
2 : L. per off sky ana. 

A remarkable feature is the presence in some cones of a number of megasporophylls 
with three ovules. I have not been able to obtain sufficient fresh material of such 
sporophylls to determine the original position of the third ovule. It appears to be 
inserted above and between the other two, though rather to one side, but is probably 
lateral in ontogeny as in the Cycadales generally. The pluriovulate condition would 
appear to be a primitive feature, and indeed the unspecialized sporophylls suggest that 
Lepidozamia is relatively primitive in the Zamiaceae. Chamberlain (1909, p. 410) has 
recorded up to five or six abortive ovules in a cone of Dioon spinulosum and, as a 
rarity, as many as four ovules in Zamia and Oeratozamia. In Lepidozamia, however, 
the condition is more common and all three ovules may develop into seeds. 

In very young seedlings the pinnae arise from the margins of the upper surface of 
the rhachis, as in Macrozamia. In successive fronds these margins are found to be closer 
together (relative to the increasing diameter of the rhachis) until the apparently 
median insertion of the adult stage is attained. This, of course, appears to be a 
specialization. The fronds superficially somewhat resemble those of Ceratozamia, but 
the sporophylls are very different. 

Regel established his genus on vegetative material alone, and maintained it later 
(1876) when he had knowledge of the cones. In this he showed greater discernment 
than other cycadologists, though Miquel was aware that it was a rather distinctive 



BY L. A. S. JOHNSON. 85 

group and, of course, Hill recognized it as Catakidozamia. The name Lepidozamia 
refers to the scale-like frond-bases clothing the stem. 

Schuster (1932) inexplicably (unless his arrangement was based, as the meanings 
of his Greek-Latin hybrid names suggest, not on morphology but on geography) placed 
his section Monoorientales (= Lepidozamia) between his other two sections of 
Macrozamia, although the latter contain extremely closely allied species. 

Key to the Species. 
1. Broadest pinnae 17-30 mm. broad, 20-40 cm. long, 17-30-nerved. N. Queensland 

1. L. hopei. 

1.* Broadest pinnae 7-14 mm. broad, 10-32 cm. long, 7-14-nerved S. Queensland and N.S.W. 

2. L. peroffsTcyana. 

1. L. HOPEI Regel* in Gartenfl. (1876), 6, and in Acta Hort. Petrop., 4 (1876), 296. 

Typiflcation: Based on a living plant cultivated by Haage and Schmidt, doubtless 
originally from Hill. The brief description unmistakably refers to this species. 

Synonymy : 

Catakidozamia hopei W. Hill in Gard. Chron. (Nov., 1865), 1107 ("from tropical 
Eastern Australia"). 

Macrozamia hopei W. Hill ex F. M. Bail., 8yn. Queensl. Fl., Suppt. 1 (1886), 52 
(from "Daintree and Johnstone Rivers". To be considered as a new publication; Bailey 
did not cite Hill's earlier binomial or description, though he attributed his new binomial 
to Hill), and Queensl. Fl., 5 (1902), 1506; also W. Hill ex C. Moore in Journ. Roy. Soc. 
N.S.W., 17 (1884), 116, nomen nudum. 

M. denisonii C. Moore & P. Muell. var. hopei (W. Hill) Schuster, Pflanzenr., IV, 
i (1932), 101 [this may be taken as truly based on Hill's name of 1865, which is the 
earliest cited]. 

Trunk 2-20 m. tall. Fronds numerous, at first suberect, later spreading, 2-3 m. long 
on adult plants, more or less puberulous, especially the rhachis, but glabrescent with 
age; petiole 30-60 cm. long (the short, much swollen, shortly tomentose base ca. 3-4 era. 
long), rounded to angular beneath, angled or (when dry) keeled and broadly two- 
furrowed above, or sometimes very slender and laterally compressed (when dry) ; 
rhachis more or less rounded when living but when dry angular beneath, more or less 
laterally compressed, angled to the median pinna-bearing keel above. Pinnae 160-200 
or more, spreading but recurved-drooping towards the ends, entire, shining above, 
recurved-falcate, broadly strap-shaped, 20-40 cm. long (lower somewhat shorter than 
the median ones, but never very short and spinescent), 15-30 mm. broad, with 15-30 
scarcely raised nerves beneath, tapered to the acute apex, slightly contracted at the 
base, with stomata on the undersurface only. Cones subsessile, usually solitary, axillary 
but often appearing quasi-terminal, surrounded at the base by several rows of fleshy, 
velvety-tomentose, subulate-tipped cataphylls. Male cones not seen, doubtless similar to 
those of L. peroffskyana. Female cones ovoid, sometimes somewhat contracted at the 
base, 40-60 (or more?) cm. long, 20-25 (-30?) cm. diam.; sporophylls ca. 5-8 cm. long, 
3-5-6-5 cm. broad, the shortly tomentose end deflexed, but the extreme tip often again 
upturned. Seeds 4-5-5 (-6?) cm. long, 2-5-3-5 cm. thick, outer coat fleshy, bright red 
when ripe. 

Distribution: North-east Queensland from the Daintree River (or further north?) 
to the Rockingham Bay region in hilly country near or within rainforest. 

Specimens examined: Babinda Creek, P. R. Messmer, 12.viii.l954 (NSW.30536); 
Bellenden Ker, C. T. White 1295, iii.1922 (NSW.40971) ; Cardwell, H. L. White, vii.1911 
(NSW, cone material only). 

* This is not nomenclaturally based on Catakidozamia hopei W. Hill, -the publication of 
^vhich appears to have been unknown to Regel. Regel cited "Katakidozamia Hopei h. Haage 
et Schm.", a garden name of no validity, though, of course, derived ultimately through 
horticultural channels from Hill. Regel's Gartenflora reference is the earlier of the two ; it 
was cited in the other 1876 paper. 



S6 THE FAMILIES OF CYCADS AND THE ZAMIACEAE OF AUSTRALIA, 

L. hopei, one of the tallest of all cycads, is reputed to reach a height of up to 20 m. 
It is very similar to L. peroffskyana in most respects, but is readily distinguished by its 
broad pinnae. Cookson (1953) describes clear differences in details of epidermal anatomy 
l)etween the two species. I have not seen sufficient cone material to determine whether 
there are significant differences in these organs. If so, they are certainly not very 
^reat, and Bailey's (1902) key is not reliable. In Bentham's treatment (1873, pp. 253-4) 
the species is not distinguished from "M. peroivskiand" , but since Bailey's time it has 
been generally recognized as specifically distinct, except by Schuster. 

Although L. hopei is at present confined to the wet tropics and is replaced further 
south by L. peroffskyana, a plant very similar to the former grew in Victoria during 
the early Tertiary. This is the following: 

Lepidozamia hopeites (Cookson) L. Johnson, comb. nov. (sp. foss. ). 

Basionym: Macrozamia hopeites Cookson in Phytomorphology, 3 (1953), 307, f. 1-5. 

Typification: "collected from a sandy bed exposed during 1947 and 1948 on top of 
the early Tertiary brown coal at the Lucifer Mine, Bacchus Marsh, Victoria". Fragments 
of pinnae. Nat. Mus. Vic. P15781 is the Holotype. 

Cookson's description and beautifully clear photographs of the epidermal anatomy 
of these pieces show the unmistakable resemblance to L. hopei. The epidermis of L. 
peroffskyana differs from these in detail, and that of Macrozamia is very different in 
orientation and shape of the cells. Dr. Cookson described the species as a Macrozamia 
pending completion of the present revision (Cookson, I.e., p. 311), but was aware at 
the time of the considerable differences between Lepidozamia and Macrozamia, which 
her studies have indeed established more firmly. 

The former occurrence of this plant, rather than one resembling L. peroffskyana, in 
Victoria illustrates the risk involved in deducing past history from present distribution. 
It may be, of course, that whereas little change has taken place in the tropical segment 
of a former single population, the more southern segment has changed considerably, 
giving rise to L. peroffskyana. On the other hand, the ancestors of the latter may have 
been already differentiated in the early Tertiary, but somewhere else. 

2. L. peroffskyana Regel in Bull. 8oc. Imp. Nat. Mose. (1857), n.l, 184, t.IV, f.20, 21, and 
in Gartenfl., 6 (1857*), 11, t.l86, f.23, 31, ibid. (1870), 227, t.660, (1875), 42 and (1876), 
4, also in Acta Hort. Petrop., 4 (1876), 295; Miq., Prodr. Syst. Cycad. (1861), 10, 22; 
A.DC, Prodr., 16, pt. 2 (1868), 547 (under Genus duUum). 

Typification: The Holotype was a living plant cultivated in the Botanic Garden, St. 
Petersburg. From the description and figures there is no doubt of its identity. 

Synonymy : 

Macrozamia denisonii C. Moore & F. Muell. in F. Muell., Fragm. Phytogr. Austral., 
1 (1858), 41 [of the three syntypes the following may be taken as Lectotype: "In 
vicinia fluminis Manning. Stephenson." NSW.40975, an old collection from Moore's 
time, but labelled only "Manning River", is almost certainly part of this. Moore (1884, 
p. 119) gives 1855 as the date of this collection]. 

Further synonymy is given by Johnson in Anderson, Flora of N.S.W., part 1 (in 
press). 

Description and Specimens: See Johnson in Anderson, I.e. 

Distribution: Subtropical eastern Queensland and North Coast of New South Wales 
to the Manning River district, usually in hilly country in wet sclerophyll forest some- 
times bordering on rainforest. 

L. peroffskyana-f has been most generally known, especially in Queensland, as 
Macrozamia denisonii, but Regel's name has clear priority and undoubtedly belongs to 

* Later than the first-cited reference. I have not seen the four Gartenflora references. 
They are quoted from Kegel's second 1876 paper. 

t The epithet has been spelt in various ways. Kegel named it in honour of Count 
Peroffsliy, an Imperial Kussian Minister and benefactor of the St. Petersburg- Botanic Garden. 



BY L. A. S. JOHNSON. 87 

the species. The distribution is scattered, small communities being found here and there 
in forested hilly country of the subtropical east coast. The plants are handsome and 
striking, but not as large as those of L. hopei; their maximum height appears to be 
about 7 m. Absurdly exaggerated claims of great individual age (to 10,000 years 
or more) for these plants have been made at tourist centres and in the popular Press. 
These estimates are usually attributed to C. J. Chamberlain, but I have found no such 
claim in any of Chamberlain's publications. His estimates of the ages of other cyead 
species, though perhaps a little high, are certainly not of this order. Plants of L. 
per off shy ana in the Sydney Botanic Gardens have attained a trunk height of 2 m. in 
less than a century. Since the plants grow only by apical increase of a single stem, and 
produce about the same number of fronds in each new crown, even the tallest individuals 
on this reckoning would be less than 500 years old. 

Bentham (1873) included L. hopei in his concept of this species. Bailey (1902) 
distinguished the two, but used unreal differences in his key. He records L. peroffskyana 
only from south of Brisbane, but localities as far north as the Tropic of Capricorn are 
cited by Moore and Mueller, Miquel and other authors, though I have not seen this 
material. Schuster (1932) used the name M. denisonii, though citing the earlier L. 
peroffskyana in synonymy, and included L. hopei as a variety. The distinction between 
the two species of Lepidozamia seems quite clear (see above, p. 86). 

II. Maceozamia Miq. 
Miq., Monogr. Cycad. (1842), 35, also Prodr. Syst. Cycad. (1861), 8, 18, and in Arch. 
Neerl., 3 (1868), 247, and Versl. Meded. Koninkl. Akad., ser. ii, 3 (1869), 50 (excl. sect. 
Lepidozamia); A.DC, Prodr., 16, pt. 2 (1868), 535 (excl. M. denisonii); Benth., Fl. 
Austral., 6 (1873), 250 (excl. M. "peroivskiana") ; Kegel in Acta Hort. Petrop., 4 (1876), 
317; Schuster, Pflanzenr., IV, i (1932), 86 (excl. sect. Monoorientales = Lepidozamia) .* 

Typiflcation : Miquel originally accepted two supposed species in his new genus, 
which he named M. spiralis and M. fraseri. M. fraseri Miq. was very briefly described 
from imperfect material and cannot, as such, be the generic type. It is now known to 
be a synonym of the Western Australian species, M. riedlei (Fisch. ex Gaudich.) C. A. 
Gardn. The generic character was taken from material illustrated in drawings of 
Ferdinand Bauer, which was the whole basis of Miquel's description under M. spiralis. 
This material is now known also to belong to M. riedlei. However, the name M. spiralis 
(Salisb. ) Miq. is nomenclaturally based on Zamia spiralis Salisb., cited by Miquel. The 
typiflcation of this is discussed below (p. 103) under M. spiralis; it is a very different 
plant from M. riedlei. Miquel (1842, p. 37) was aware that Salisbury's plant may not 
have been the same as those he described, but saw no specimens. Under M. spiralis he 
also cited, but did not see, plants from Port Jackson previously described by R. Brown 
under "Zamia spiralis Salisb.", but which in fact belong to M. communis L. Johnson. 
Later, in Versl. Meded. Koninkl. Akad., 15 (1863), 368, Miquel pointed out that Bauer's 
figures and hence his own description of M. spiralis pertained in fact to the same species 
as his M. fraseri. He then used "M. spiralis" for M. communis (with some confusion with 
plants later described under M. miquelii (F. Muell.) A.DC). Later authors used "M. 
spiralis" chiefly for either M. communis or M. lucida L. Johnson, but the name is herein 
(p. 103) restored to Salisbury's original species. 

Hutchinson, in Eew Bull. (1924), 51, named "M. spiralis" (without author) as type 
species of Macrozamia Miq., but was unaware of the complicated circumstances and had 
made no detailed study of the genus. His selection is not binding. The International 
Code of Botanical Nomenclature (Lanjouw, 1956) lays down that the type of a generic 
name is a species (not a specimen nor, one may add, a specific name). Now a species 
consists of living organisms, past, present or future. Its nature must be elucidated by 
biological enquiry, and the circumscription of our concept of it is determined by 
taxonomic, not nomenclatural, considerations. Although the specific name Macrozamia 
spiralis (Salisb.) Miq. must be applied to a group of organisms including the plant 

* Only important treatments purporting to cover the wtiole genus are cited. 



88 THE FAMILIES OF CYCADS AND THE ZAMIACEAE OF AUSTRALIA, 

described by Salisbury, this plant was certainly only in the vaguest way part of the 
concept in Miquel's mind when he described his new genus. On the contrary, he drew 
his description almost wholly from the plants represented in Bauer's plates. 

Consequently I consider that the Type Species of Macro^amia Miq. is the species 
to which Bauer's plants belong, whatever its name is held to be; in my treatment this 
is M. riecllei (Fisch. ex Gaudich.) C. A. Gardn. 

By this typiflcation we may preserve Miquel's sectional treatment, as far as it is 
applicable. True M. spiralis (Salisb.) Miq. belongs to section Parazamia (Miq.) A.DC. 
(based on Encephalartos sect. Parazamia Miq., typified by M. pauli-guilielmi W. Hill and 
F. Muell.), while M. riedlei belongs to the section regarded by Miquel as typical: 
section Macroizamia of my treatment (Encephalartos sect. Macrozamia Miq., Macrozamia 
sect. Genuinae Miq.). If M. spiralis as fixed by Salisbury's type were taken as type 
species, Miquel's "typical" section would require a new name and his Parazamia would 
become Macrozamia. 

It is regrettable that nomenclatural procedure demands so detailed an argument 
to settle permanently (one may hope) a matter of plain common sense. 

Synonymy: There are no actual generic synonyms of Macrozamia when Lepidozamia 
Regel is excluded. Infrageneric names are listed below, under the two sections. The 
genus was included in Encephalartos by F. Mueller intermittently over many years (see 
refs. under the species), also briefly by Miquel, in Yersl. Meded. Koninkl. Akad., 15 
(1863), 368. Before 1842, individual species were referred to Zamia or Encephalartos, 
and Mueller, Fragm. Phytogr. Austral., 5 (1866), 172, stated: (translation) "I have 
reduced all the species of Encephalartos in the collections of the Melbourne museum 
subgenerically to Zamiae." Despite this remark, he had continued to use Encephalartos 
(incl. Macrozamia) on the same page, and certainly did not formally establish 
Encephalartos in subgeneric rank under Zamia, as Schuster's (1932, p. 86) citation 
seems to indicate. In 1881 in his paper on M. moorei (q.v.) Mueller again expressed 
his view that Macrozamia was a "sub-generic group" of Encephalartos, if indeed both 
were not to be placed under Zamia, adding that "all genera are mere artificial groups to 
facilitate classification, and aid memory, while species in their true sense are originally 
created beings, which when perished . . . would require the godly might as much for 
their restoration as they did for their origination". This philosophy should be borne 
in mind when interpreting any pronouncement of Mueller's on classification or nomen- 
clature. 

More or less palm-like plants with a usually unbranched stem forming a sub- 
terranean caudex or a rather massive aerial trunk, clothed by the persistent leaf-bases, 
all parts more or less pubescent when very young, but glabrous (except petiole-bases) 
at maturity. Fronds few to numerous, simply pinnate (but the pinnae sometimes 
dichotomously divided), rhachis straight or twisted. Cataphylls present, angular- 
subulate, at first silky or woolly, finally often glabrescent. Base of petiole expanded, in 
most species silky or woolly-tomentose. Pinnae numerous in mature plants, spreading 
or secund, inserted near the edges of the rhachis towards the adaxial side, simple or 
once to thrice forked, straight or falcate, contracted and sometimes callously thickened 
at the base, stomata on lower or both surfaces. Cones of both sexes stalked, axillary 
among the fronds (with several cataphylls below and sometimes on the stalk), the 
spirally-arranged sporophylls glabrate (often glaucous) and spiniferous at the ends 
(spines sometimes almost obsolete on lower sporophylls). Male sporophylls with a 
broadly cuneate fertile region bearing the sporangia beneath in two separate to more 
or less confluent areas, and an upturned end terminated by an erect or spreading-erect 
spine, the latter sometimes very short on the lower sporophylls. Female sporophylls 
appearing peltate, with an angular-rounded stipes and a laterally expanded more or less 
terminally compressed end, the latter with a narrow transverse wing (continuation of 
the mai-gins) terminated by a more or less erect short to long spine; the two ovules 
borne on the inward-facing margins ("inner surface") of the sporophyll-end; the 
sporophylls falling with the seeds attached by the outer fleshy red or yellow seed-coat, 
the inner seed-coat hard. 



BY L. A. S. JOHNSON. 89 

Chromosome number: x = 9. The number 2w = 18 has been found in M. miquelii, 
M. moorei and material named "M. tridentata" (refs. in Darlington and Wylie, 1955), and 
also in M. communis (Brough and Taylor, 1940, as "ilf. spiralis"). Light (1924) records 
n - 12 for M. riedlei (as M. fraseri), but this early count is doubtful; if correct it 
suggests triploidy. 

Endemic in Australia, with 14 species in the warm-temperate and subtropical 
regions, barely reaching the Tropic: 12 of them in eastern Australia (New South Wales 
11, Queensland 5), 1 in Central Australia, 1 in south-western Australia. This exclude3 
the quite distinct Lepidozamia (see above, p. 83), which most authors have included in 
the genus. The species grow mostly on nutritionally poor, frequently siliceous, soils in 
sclerophyll communities, in association with members of the old "Australian" floristic 
element. Most parts of the plants contain a toxic substance, macrozamin (Hurst, 1942; 
Webb, 1948; Gardner and Bennetts, 1956), and various species are held responsible for 
a form of stock poisoning known as "wobbles" (or miscalled "rickets"), especially com- 
mon in cattle which can eat the tough fronds. The aborigines ate the starchy endosperm 
of the seeds after roasting and prolonged washing to remove the poison (Bailey, 1902; 
and various notes in herb. NSW). Several species have been cultivated as ornamentals. 

As indicated above (p. 72), the taxonomy and especially the nomenclature of the 
genus have been quite extraordinarily confused and unstable. 

Key to the Sections and Species. 

Note : Juvenile plant.s of Macrozamia may differ considerably from the adult forms in 

details of the fronds ; in particular the pinnae are usually toothed at the ends and the 

petioles are long and slender. The stomata of juvenile plants are confined to the lower 

pinna-surface. Only adult organs are described in the key and descriptions given here. 

Similarly, cone dimensions apply to mature cones (males at sporangial dehiscence, females 

at ripening of the seeds). Petiole lengths do not include the woolly expanded hase and 

may be taken as the distance from the end' of the tomentose portion to the lowest pinnae. 

The stomata are readily seen at a magnification of xl5-20 diameters, or with i^ractice at 

xlO diameters, especially in fresh material. To determine their presence or absence the 

upper pinna-surface should be compared with the lower, which, of course, always bears stomata. 

1. Large plants with 15-150 fronds in the crown, aerial trunk present or absent, caudex 

15-100 cm. diam. Fronds 50-300 cm. long ; rhachis not strongly twisted, usually angled 

beneath (at least in petiolar portion) when dry, flat to angled and laterally 2-channelled 

(when dry) above, 6-30 mm. broad at lowest pinnae. Pinnae straight for most of their 

length, spreading, nerves visible but narrow and not or scarcely raised on the lower 

surface ivhen dry; bases markedly callous on the anterior margin (± rugose when dry> ; 

the lowest pinnae usually progressively reduced and sometimes spine-like (not always so 

in M. miquelii). Mucilage canals present in the pinnae. $ cones 15-90 cm. long, S-27 cm. 

diam., longest sporophyll spines 1-10 cm. long; seeds 2-5-8 cm. long, d" cones 15-50 (-60?) 

cm. long, longest sporophyll spines 1-5-5 cm. long i. Section Macrozamia. 

2. Reduced spine-like pinnae extending almost to the base of the frond (spine-free petiole, 
excluding swollen base, 0-10 cm. long). Plants with massive trunks to 2-5 m. tall, 
60-80 cm. diam. Fronds 150-300 cm. long with 150-250 pinnae. Rhachis 12-30 mm. broad 
at lowest pinnae. Stomata on both surfaces of pinnae. ? cones 40-90 cm. long, d cones 

30-45 cm. long. C. Queensland and N.S.W. Far N. Coast 1. M. moorei. 

2.* Reduced pinnae not extending to near base of frond (petiole 12-60 cm.). Aerial trunk 
present or absent. Fronds 50-220 (-250?) cm. long, with 60-170 pinnae. Stomata on 
both surfaces of pinnae or on lower surface only. Cones various. 
3. Stomata on both upper and lower surfaces of pinnae. Pinnae simple. "West and Central 
Australian species. 

4. Fronds not or scarcely glaucous. Seeds 3-5 (-6?) cm. long, 2-3-5 cm. thick. Spines of 
uppermost $ sporophylls 1-5-6 (or more?) cm. long, lower sporophylls usually with 
short broad spines. S.W. Australia 2. M. riedlei. 

4.* Fronds markedly glaucous. Seeds 6-8 cm. long, 4-5-5 cm. thick. Spines of tippermost 
9 sporophylls ca. 1-2 cm. long, lower sporophylls almost spineless. Central Australia 

3. ilf. macdonnellU. 

3.* Either stomata on lower surface of pinnae only and pinnae simple or stoniata on both 
surfaces and most pinnae forked. Eastern species. 

5. Pinnae simple, stomata on lower surface only. 

6. Pinnae thin and rather lax, easily bent, shining above, often rather crowded along 
the rhachis. Most sporophylls of $ and d' cones short-spined ; longest spines (?) 



90 THE FAMILIES OF CYCADS AND THE ZAMIACEAE OF AUSTRALIA, 

2-3 J cm., (d") 1-2 (2^) cm. Most of the ? sporophyll-ends less than twice as broad 
as high ; spines of cf sporophylls slender, 2-5 mm. broad at the base. Seeds 2-5-3-5 
cm. long. Eastern parts of C. and S. Queensland and Far N. Coast (Richmond R. ) 

of N.S.W 4. M. miquelii. 

e.'* Pinnae thicker and more rigid, rather dull, often rather widely separated along the 
rhachis. Many sporophylls of ? and d" cones long-spined ; longest spines (?) 4-10 cm., 
(cT) 2-5 cm. Most of the ? sporophyll-ends at least twice as broad as high (excl. 
spine) ; spines of d" sporophylls 5-12 mm. broad at the base. Seeds 3-4-5 cm. long. 
Macleay River to Bega, also Goulburn River Valley, eastern N.S.W. . . 5. M. communis. 
5.* Most pinnae dichotomously once-divided. Stomata on both surfaces. Coonabarabran 

and country north of Liverpool Range, N.S.W 6. M. diplomera. 

1.* Rather small plants with 2-12 (rarely up to ca. 40 in M. lucidal) fronds in the crown, 
caudex almost or wholly subterranean, 5-30 cm. diam. (more in M. platyrachisi) . Fronds 
30-120 cm. long; rhachis straight or twisted, rounded or angular beneath, rounded to flat 
or concave above (sometimes with two lateral grooves), 3-lS mm. broad at lowest pinnae. 
' Pinnae straight or falcate, spreading or erect, nerves thick and i)rom,inent on the lowe'^ 
surface (especially when dry) ; bases not or only slightly callous on the anterior margin 
(not rugose when dry) ; the lowest pinnae usually not much reduced (sometimes somewhat 
so in spp. with secund pinnae or twisted rhachis). Mucilage canals absent from the pinnae. 
? cones (10-)15-25 cm. long, 6-10 cm. diam., longest sporophyll-spines 0-5-4 (-5?) cm. 
long; seeds 2-3-5 cm. long. d" cones 10-25 cm. long, longest sporophyll-spines 0-2-1-5 cm. 

long ii. Section Parazamia. 

7. Rhachis not or moderately spirally twisted (0-180°, rarely to 360° but if so petiole 20-40 
cm. long or pinnae divided) though sometimes recurved near the end. Pinnae simple or 
divided. 
8. Pinnae simple. Stomata on lower surface only. 
9. Broadest pinnae 15-20 mm. broad, 15-18-nerved. Rhachis flat, 13-18 mm. broad at 

lowest pinnae. C. Queensland 7. M. -platyrachis. 

9.* Broadest pinnae 3-12 mm. broad, 5-11-nerved. Rhachis rounded or ± flat or concave 

above, 3-9 mm. broad at lowest pinnae. 

10. Pinnae spreading in the living state. Rhachis scarcely to moderately (0-180°, rarely 

360°) twisted; petiole rounded or flat above, 15-50 cm. long. Broadest pinnae 6-12 

mm. broad, 12-35 cm. long, not or scarcely glaucous. 

11. Pinnae glossy, ± falcate, whitish at the base when living, the longest ones 15-35 cm. 

long, 7-12 mm. broad. Petiole rounded, (20)25-50 cm. long. S. Queensland, 

N.S.W. (Far N. Coast) 8. M. lucida. 

11.* Pinnae rather dull, straight to ± falcate, pinkish to red or orange at the base when 
living, the longest ones 12-25 cm. long, 5-10 mm. broad. Petiole rounded or 
subangular, (15)20-40 cm. long. N.S.W. (C. Coast to Goulburn R. Valley and 

Dunedoo) 9. M. spiralis. 

10 * Pinnae secund, rising ± vertically from rhachis in living state. Rhachis not or 
slightly twisted, but recurved near the end; petiole concave above, 5-22 cm. long, 
broadest pinnae 3-8 mm. broad (8-20 cm. long), often ± glaucous. N.S.W. (C.W. 
Slopes) 10. M. secunda. 

8.* Pinnae once to thrice dichotomously divided. Stomata on lower or both surfaces. N.W. 
Slopes of N.S.W. 
12. Stomata present on both surfaces of pinnae. Pinnae once or twice divided, rather 
stiff, dull above, often ± glaucous beneath. Rhachis little if at all twisted (0°-90°-180°). 

N.S.W. (Warrumbungles, Pilliga Scrub and Warialda-Howell, on sandy soil) 

11. M. heteromera. 

12.* Stomata conflned to lower surface of pinnae. Longest pinnae usually 2-3 times divided 
(if only once divided segments nevertheless very narrow), rather lax, dark green 
above, not glaucous. Rhachis usually twisted near the end (90°-180°-360°). N.S.W. 

(Nandewars and surrounding district, on stony hillsides) 12. M. stenomera. 

7.* Rhachis strongly twisted (360° or much more, rarely less in depauperate short fronds of 
M. pauli-guilielmi ssp. plurinervia but if so then the petiole 5-15 cm. long). Pinnae simple 
(but sometimes toothed). 
13. Broadest pinnae 3-7 mm. broad', 3-10-nerved, adult ones entire or closely 2-toothed at 
the apex. S. Queensland and N.S.W. (N. Tablelands, N.W. Slopes and Manning R.- 
Lake Macquarie) 13. M. pauli-fjjiilielmi. 

13.* Broadest pinnae 9-17 mm. broad, 10-13-nerved, adult ones 2-7-toothed at the apex. 
N.S.W. (N. Coast : Richmond R.-Macleay R.) 14. M. fawcettii. 

i. Section Macrozamia. 
Typiflcatiofi : As for the genus, the species now known as M. riedlei (Fisch. ex 
Gaudich.) C. A. Gardn. 



BY L. A. S. JOHNSON. 91 

Synonymy (infrageneric) : 

Encephalartos Lehm. sect. Macrozamia Miq. in Tersl. Meded. Koninkl. Akad., 15 
<1863), 368 (this included only species of the section as here understood, exclusive of 
sect. Parazamia and of Lepidozamia. Type as for the genus Macrozamia). 

Macrozamia Miq. sect. Genuinae Miq. in Arch. Neerl., 3 (1868), 247, and in Versl. 
Meded. Koninkl. Akad., ser. ii, 3 (1869), 50 (absolutely equivalent to the above). 

Macrozamia sect. Polyorientales Schuster, Pflanzenr., IV, i (1932), 87 (in part, 
excluding several species. Nomenclaturally illegitimate since it includes the type of 
the earlier section Parazamia Miq., though not based on this. I select M. moorei F. 
Muell. as Lectotype, since the other names included by Schuster are confused. Even 
should M. moorei be sectionally separated, Schuster's name remains rejected). 

Macrozamia sect. Polyorientales Schuster subsect. Attenuatae Schuster, I.e. (for 
the greater part. I select M. moorei F. Muell. as Lectotype). 

Macrozamia sect. Monooccidentales Schuster, I.e. (this comprises only M. riedlei and 
M. macdonnellii, considered conspecific by Schuster, and distinguished sectionally by 
no tangible character from his sect. Polyorientales. It is nomenclaturally illegitimate 
since it is based on the type of the genus — see above, p. 87. Schuster nowhere refers 
to Miquel's earlier sectional names). 

Pinnae predominantly straight and spreading, nerves visible but narrow and not 
or scarcely prominent on the lower surface when dry, apices always entire in adult 
fronds, bases markedly callous and (when dry) rugose on the anterior margin, mucilage 
canals present in the pinnae, the lowest pinnae usually progressively reduced. 

Usually large plants with many large fronds, aerial trunk sometimes present and 
robust. Rhachis never strongly twisted (except in rare aberrant individuals), usually 
flattened above and angled beneath when dry. Cones usually large (females 15-90 cm. 
long, 8-30 cm. diam.; males 15-45 cm. long), sporophyll-spines short to long, always 
erect. 

This is the more widespread section, having one species each in the south-west 
and the centre of the continent, as well as four in the east. The species show less 
evidence of reduction than those of sect. Parazamia, which suggests that sect. 
Macrozamia is the more primitive, or more conservative, of the two. Individuals of 
the various species, unlike those of section Parazamia, usually occur plentifully in 
fairly dense communities and may form quite an important and conspicuous part of 
the vegetation. Regeneration is vigorous in most places. In distribution the species 
show a replacement pattern without overlap, except in the case of M. moorei and 
M. miquelii, where the former is found both north-west and south of the latter's area, 
but not within it. Naturally-occurring hybrids are consequently unknown within the 
section. On the other hand the distribution of species of sect. Macrozamia frequently 
overlaps that of species of sect. Parazamia, and mixed stands occur in places. A few 
hybrids may be found, but the specific limits do not break down. 

The six species of this section fall naturally into the following three groups: (a) the 
remarkable M. moorei alone, (&) the two vicarious western and central species M. 
riedlei and M. macdonnellii, (c) the three vicarious eastern species M. miquelii, M. 
communis and M. diplomera. Groups (h) and (c) may themselves be considered as a 
pair of vicarious "superspecies" (Text-flgure 2). 

1. M. MOOEEi F. Muell. in Chemist and Druggist, Australas. Suppt., 4 (March, 1881), 84. 
Tyjnfication: "With certainty known from the mountainous regions of Queensland 
at the verge of the tropics", with a reference to its cultivation by Charles Moore in 
the Sydney Botanic Gardens, is the only information in the original publication, but 
in August, 1881 (see below) Mueller cited "In collibus praesertim altioribus lapidosis 
ad originem fluminis Nogoa-River imprimis circum urbem [sic!] Springsure satis 
frequens; J. G. Macdonald; P. A. O'Shanesy". (I have not seen these Syntypes and 
therefore cannot name a lectotype, but the identity of the species is unmistakable 
from the description and locality.) 



92 THE FAMILIES OF CYCADS AND THE ZAMIACEAE OF AUSTRALIA, 

Synonymy : 

Encephalartos moorei (P. Muell.) F. Muell., Fragm. Phytogr. Austral., 11 (August, 
1881), 125. In the preamble to his original publication of March, 1881, Mueller 
vacillated, as he had done for many years in various publications, regarding the generic 
recognition of Macrozamia as distinct from Encejilialartos. The title of his paper 
referred to a "species of Encephalartos" , and in his discussion he appeared to imply that 
Macrozamia was to be regarded as a subgenus of Encephalartos, if indeed both of 
these were not to be included in Zamial However, although the preamble was equivocal, 
in the formal treatment the species was described under the binomial Macrozamia 
moorei, and in his later publication in August of the same year, when placing 'he 
species definitely under Encephalartos, Mueller cited Macrozamia moorei as a synonym, 



Text-fig. 2. — Distribution of Macrozamia section Macrozamia : 1 : M. moorei ; 2 : M. riedlei ; 
3: M. macdonnellii ; 4: M. miquelii; 5: M. comnmnis ; 6: M. diplomera. 

with full reference to its earlier publication. Clearly, Mueller did not really know 
whether or not he himself accepted his binomial Macrozamia moorei at the time of 
publication, and its validity under the I.C.B.N. depends on this. Since the case may be 
argued either way, I consider that M. moorei F. Muell. should be accepted as validly 
and legitimately published in March, 1881, thus preserving the established nomenclature 
and citation. The alternative is to reject M. moorei F. Muell., March, 1881, as invalid; 
the first valid publication would then be as Encephalartos moorei F. Muell., August, 
1881, and the first valid publication under Macroza7nia would be as M. moorei F. Muell. 
ex F. M. Bail., Syn. Queensl. Flora (1883), 501, which is not based on a reference to 
Encephalartos moorei F. Muell., but is to be treated as a new publication. A much 
more detailed nomenclatural discussion of this matter is possible, but, being devoid of 
biological significance, would be unprofitable. For further citations and misapplications 
see Johnson in Anderson, Flora of N.S.W., part 1 (in press). 

Description, specimens and further discussion: See Johnson in Anderson, I.e. 

Chromosome number: 2n = 18. 

Distribution: Queensland and New South Wales: Two disjunct areas, in Central 
Queensland (Springsure-Carnarvon Range districts) and the Clarence River 
(Dalmorton) district of the North Coast of N. S. Wales, in dry sclerophyll forest or 



BY L. A. S. JOHNSON. 93 

in the ecotone between semi-wet sclerophyll forest and rainforest, on rather shallow, 
rocky or stony soils. 

Apparently an old species now fragmented into two widely disjunct groups of 
populations, but within these still vigorous, M. moorei stands somewhat apart from 
the other species of the section, particularly in the long series of reduced pinnae 
extending almost to the frond-base, and in the large numbers of cones (especially 
males). Chamberlain (1913) was so impressed by the numerous lateral cones that he 
devoted a whole paper to this species and suggested a close relationship with the 
Mesozoic Bennettitales. However, the nature of the "strobilus" and its parts differs so 
greatly in Cycadales and Bennettitales that the relationship is now held to be rather 
a distant one. Certainly Macrozamia cannot be derived from anything like the 
Bennettitales. 

The plants are massive and spectacular and of fairly rapid growth; individuals 
with trunks 2 m. tall and 60-70 cm. thick in the Sydney Botanic Gardens are less than 
100 years old. Some apparent natural hybrids between M. moorei and M. lucida (of 
section Parazamia) are discussed in the forthcoming Flora of N.8.W., part 1. 

2. M. EiEDLEi* (Fisch. ex Gaudich.) C. A. Gardn., Enum. PI. Austral. Occid. (1930), 3; 

C. A. Gardn. in Gardn. and Bennetts, Toxio PI. W. Austral. (1956), 6 (s. ampl.). 

Basionym: Cycas riedlei* Fisch. ex Gaudich. in Freycinet, Voy. autour du Monde, 
Bat. (1826), 434. 

Typiftcation: "In Novae Hollandiae ora occidentali (Port du Roi George)." This 
is in herb. Paris. I have not seen it, but it was seen by Miquel, and from the locality 
there is no doubt of its identity. 

SynonyTny : 

Macrozamia fraseri Miq., Monogr. Cycad. (1842), 37 [Schuster (1932) states that 
this is "sine descr.", but Miquel's description, brief as it is, is sufficient to validate the 
name]; Miq., Prodr. Syst. Cycad. (1861), 8, also in Arch. Neerl., 3 (1868), 247, and in 
Versl. Meded. Koninkl. Akad., ser. ii, 3 (1869), 50; Benth., Fl. Austral., 6 (1873), 252. 
Not M. spiralis (Salisb. ) Miq. [var.] S fraseri Regel in Acta Hort. Petrop., 4 (1876), 
318. 

M. preissii Lehm., Pugill., 8 (1844), 31, descr. (not seen) and in Cat. Hort. Hamb. 
(1842), nomen? (not seen); Heinzel in Nov. Act. Acad., 21, i (1844), 203, t. 10-13 (not 
seen); Lehm., PI. Preiss., 1 (1845), 645; F. Muell., Fragm. Phytogr. Austral., 1 (1858), 
41, and various later references; A.DC, Prodr., 16, pt. 2 (1868), 525; Regel in Acta Hort. 
Petrop., 4 (1876), 318; Schuster, Pflanzenr., IV, i (1932), 101 (in part, excl. syn. 
M. macdonnellii and Central Australian citations). 

Encephalartos preissii (Lehm.) F. Muell. in Quart. Journ. Pharm. Soc. Vict., 2 
(1859), 90 [this is validly published, but in the further reference given by Schuster 
(1932) to Miquel (1863) the name is mentioned only in synonymy of E. fraseri}. 

E. fraseri (Miq.) Miq. in Versl. Meded. Koninkl. Akad., 15 (1863), 368. 

E. oldfieldii Miq., I.e., 370. 

Macrozamia oldfieldii (Miq.) A.DC, Prodr., 16, pt. 2 (1868), 535; Miq. in Arch. 
N6erl., 3 (1868), 250 [cited by Schuster from the separate as "Nouv. Mater. (1868) 58"]; 
Miq. in Versl. Meded. Koninkl. Akad., ser. ii, 3 (1869), 53. [Schuster (1932) gives this 
the first citation, as "Nieuw. Bijdr. Cycad. (1868) 53", but the paper thus entitled was 
published in the above-cited journal, the title-page of which is dated 1869; I can find no 
indication in it of an earlier date of publication of the paper, though this is possible. 
It matters little.] 

* In publishing the combination, and consistently since, Gardner, who apparently did not 
consult the original, writes "reidlei" (all epithets here quoted with lower case initials though 
originally capitalized). Miquel (1868, 1869) and" Schuster (1932) both cited "riedleyi" ; 
and Regel (1876) gave "riedeli". The species was named for the French gardener Riedl6, 
and was correctly spelt "riedlei" by Gaudichaud. Though he used Fischer's herbarium name, 
Gaudichaud suspected that the plant's affinity was with Zamia rather than Cycas. The 
description is brief but sufficient. 



94 . THE FAMILIES OF CYCADS AND THE ZAMIACEAE OF AUSTRALIA, 

Encephalartos dyeri* P. Muell. in Chemist and Druggist, Australas. Suppt., 8 
(1885), 12. 

Uacrozamia dyeri* (P. Muell.) C. A. Gardn., Enutn. PI. Austral. Occid. (1930), 3. 

M. preissii Lehm. ssp. dyeri^ (P. Muell.) Schuster, Pflanzenr., IV, i (1932), 102. 

Caudex sometimes mostly subterranean, but frequently forming a trunk 1-5 m. tall 
or said to be sometimes procumbent, 60-120 cm. diam. Pronds numerous (usually 50-100 
or more?), at first erect, later spreading or drooping, up to 150-200 cm. long on mature 
plants; spine-free petiole (excluding the silky-woolly swollen base) 12-30 (or more?) 
cm. long, rhachis not markedly twisted, more or less flattened, 10-20 mm. broad at the 
lowest pinnae, flat to somewhat concave or convex above, often somewhat keeled 
distally, with two narrow lateral grooves (more marked when dry) decurrent from 
the bases of the pinnae, convex and usually subangular to angular beneath. Pinnae 
100-150 (or more?), spreading, but the two ranks often inclined to each other, 
forwardly directed at an acute angle (the lowest least so), mostly rather crowded, but 
the lowest ones more distant, decidedly rigid, entire, straight, linear, the longest 20-35 
cm. long, several of the lowest progressively reduced and spine-like, 6-11 (-15 in dyeri 
forms) mm. broad with 8-15 (-18 in dyeri) scarcely raised nerves beneath, gradually 
tapered to the pungent apex, contracted to the pale or somewhat reddish and anteriorly 
callous and (when dry) rugose base, green but not shining, with numerous stomata on 
both upper and lower surfaces. Cones stalked, in either sex l-several(?) per plant, 
axillary among the fronds,* the base of the stalk surrounded by several spine-like, 
angular-subulate to strap-shaped cataphylls (pubescent or tomentose when young), 
sometimes with a few shorter decurrent cataphylls on the stalk itself. Male cones 
cylindrical, often somewhat curved when old, 20-40 (-60 in dyeri) cm. long, ca. 10 cm. 
diam. (sometimes more?); sporophylls cuneate to obovate-lanceolate, 3-5 (-6 ace. to 
Schuster) cm. long, ca. 1-5-2-5 cm. broad, the triangular to elongate upturned spines 
from almost obsolete to 1-5-5-5 cm. long (the longest near the apex of the cone). 
Female cones ovoid-cylindrical, 25-45 (or more?) cm. long, 15(?)-25 cm. diam.. stalk ca. 
12-20 (or more?) cm. long, to 3-4 (-5?) cm. thick; sporophylls ca. 5-8 (-10 ace. to 
Schuster) cm. long (stipes ca. 3-6 cm.), the expanded ends glaucous, 4-5-9 cm. wide, ca. 
3-5 cm. high, the flattened erect spines increasing in length towards the apex of the 
cone, the lowest from almost obsolete to ca. 0-8 cm. and often lacerate, the uppermost 
from 1-5-6 (-10 ace. to Schuster, but doubtful) cm. long, broad-based; inner parts of 
the cone salmon-pink when fresh. Seeds 3-5 (-6?) cm. long, 2-5-3-5 cm. thick, outer 
coat bright red when ripe. 

Chromosome numter: n = 12, according to Light (1924), but this may well be a 
miscount, since it does not agree with related species {n = 9). 

Distribution: Western Australia: South-western and southern coast region, from 
the Murchison River to east of Esperance, in dry and wet sclerophyll forest (or scrub) 
on sandy and lateritic soils. 

Specimens examined: Western Australia: Bullsbrook area, 27 miles N.N.E. of 
Perth, Kennington, 31.V.1957 (NSW.42034) ; Armadale, J. M. Griffiths, viii.1900 
(NSW.40660); Big Brook, M. Koch 1419, vi.l9— (NSW.40662); Bow River, S. W. 
Jackson, xii.1912 (NSW.40661); Western Australia, J. B. Cleland, 1907 (NSW.40663); 
Western Australia, E. Brown, 1898 (NSW.40656). 

M. riedlei exhibits considerable variation in size and habit, but if Gardner (in 
Gardner and Bennetts, 1956) is correct, all the Western Australian populations are to 
be regarded as conspecific. The only authentic material of M. dyeri which I have seen 
is a pair of collodion cuticular pulls from dorsal and ventral surfaces of a pinna of 
an apparent isotype ("Esperance Bay, W.A.") in herb. Kew, made available by the 

* Included in the synonymy on the testimony of Gardner (1956). See discussion below. 

t Gardner (1956) states that "the female has a solitary terminal central cone". Doubtless 
this is based on superficial observation; whatever the species, the cone (in individuals bearing 
but one) may appear terminal but on careful inspection is seen to be lateral. Again, it 
would be surprising if 31. riedlei never bore more than a single cone, unlike its congeners. 



BY L. A. S. JOHNSON. 30 

kindness of Dr. R. Melville. These show stomata on both surfaces. The descriptions 
of M. dyeri, however, indicate that the pinnae are considerably broader than any which 
I have seen in available specimens of M. riedlei. Miss A. Baird (pers. comm.) indicates 
that M. riedlei varies greatly in stature in various parts of Western Australia and, as 
in M. communis in New South Wales, the development of the trunk above ground level 
tends to be greatest on shallow soils, in which the caudex cannot be pulled far belov/ 
ground by the contractile roots. Gardner (I.e.) states that the largest forms occur 
towards the northern and eastern limits of its distribution. 

While lacking personal field experience and sufficient herbarium materials of the 
Western Australian populations, I must follow Gardner in recognizing only a single 
species, though with reservations. 

Both R. Brown (1810) and Miguel (1842) included material of this species under 
the names Zamia spiralis /Macrozamia spiralis. Bauer's figures reproduced by Miquel 
(I.e., PI. 4-5) actually represent M. riedlei, as Miquel himself (1863) later pointed out. 
This is discussed above under the typification of the genus, and below under M. spiralis. 
Schuster (1932) included M. maedonnellii under his "typical" M. preissii (= M. riedlei), 
though he gave M. dyeri subspecific rank. Bentham (1873) likewise included M. 
maedonnellii under M. fraseri (= M. riedlei). However, M. maedonnellii seems reason- 
ably distinct and is certainly isolated. 

Fronds of M. riedlei often appear very similar to those of M. communis, but the 
pinnae are generally even more rigid and may be distinguished by the presence of 
stomata on the upper surface. 

3. M. MACDONNELLH* (F. Muell. ex Miq.) A.DC, Prodr., 16, pt. 2 (1868), 537; Miq. in 
Arch. N^erl, 3 (1868), 249, and in FersL Meded. Koninkl. Akad., ser. ii, 3 (1869), 
53; F. Muell., Fragm. Phytogr. Austral, 9 (1875), 124 (and ibid., 2 (1861), 179, 
nomen) ; Ewart and Davies, Fl. Northern Territory (1917), 19. 

Basionym: Encephalartos maedonnellii* F. Muell. ex Miq. in Yersl. Meded. Koninkl. 
Akad., 15 (1863), 376. 

Typification: "Nova Hollandia centralis, ad flumen Neales in Macdonnell-range, 
unde reportavit eel. peregrinator J. M. Stuart" (not seen, but of certain identity from 
the general locality and description. The present Neales River is a rarely-fiowing 
stream in South Australia, debouching into Lake Eyre. There are certainly no 
Macrozamia species in this most arid region. Either there was some confusion as to 
the precise locality, or it was merely meant that the plant was collected at some time 
on the Neales River — Macdonnell Range stage in Stuart's epic expedition. The plant 
could not be missed in the Macdonnells). 

Trunk usually developed above ground (the plants grow on shallow soils), 1-2 
(-3) m. tall, but often more or less procumbent, 60-80 (or more?) cm. diam. Fronds 
numerous (50-100 or more), at first erect, later spreading or drooping, 150-220 cm. 
long on mature plants; spine-free petiole (excluding the silky-woolly expanded base) 
12-25 (or more) cm. long; rhachis not markedly twisted, more or less flattened, 15-25 
mm. broad at the lowest pinnae, flat to somewhat concave but usually more or less 
keeled above, with two narrow lateral grooves (more marked when dry) decurrent from 
the bases of the pinnae, convex and subangular to angular (when dry) beneath. Pinnae 
120-170, spreading but the two ranks usually inclined to each other, forwardly directed 
at an acute angle, mostly rather crowded but the lowest usually 4-5 cm. apart, decidedly 
rigid, entire, straight, linear, the longest 20-30 cm. long, ca. 8-20 of the lowest pro- 
gressively reduced and spine-like, 7-11 mm. broad with 8-15 scarcely-raised nerves 

* By Miquel (1863, 1868, 1869), but not by Mueller or De CandoUe, spelt "macdonelli" 
but the single "n" is to be regarded as an unintentional error. Miquel correctly spelt 
"Macdonnell-range". F. Mueller (1875) Is usually cited as the author of the valid combination 
under Macrozamia, but De Candolle definitely made it in 1868, though as a species incerta. 
De CandoUe's treatment was published in mid-July, 1868, according to Stearn (1941) ; 
Miquel's paper in Archives Neerlandaises (1868) was probably published late in that year, 
although no month is indicated on the cover or title page of the part. 



90 THE FAMILIES OF CYCADS AND THE ZAMIACEAE OF AUSTRALIA, 

beneath, gradually tapered to the pungent apex, contracted to the pale anteriorly callous 
and (when dry) often somewhat rugose base, dull and (rhachls also) decidedly 
glaucous on both surfaces (especially when living), with numerous stomata on both 
upper and lower surfaces. Cones stalked, in either sex 1-several per plant, axillary 
among the fronds, the base of the stalk surrounded by several spine-like cataphylls, 
usually with a few shorter decurrent cataphylls on the stalk itself. Male cones 
cylindrical, often somewhat curved when old, 25-40 cm. long, ca. 8-10 cm. diam.; 
sporophylls cuneate-obovate, 3-4 cm. long, ca. 1-5-2 cm. broad, the expanded ends very 
glaucous and the triangular-acuminate to elongate upturned spines from obsolete on 
the lower sporophylls to 1-5-2-5 cm. long near the apex of the cone. Female cones 
ovoid-cylindrical, 40-50 cm. long, 20-27 cm. diam., stalk 12-20 (or more?) cm. long, to 
4-5 cm. thick; sporophylls 7-10 cm. long (stipes ca. 5-7 cm.), the expanded ends 
glaucous, 8-12-5 cm. wide, ca. 4-6 cm. high, the spines almost or quite obsolete on the 
lower sporophylls, flattened and triangular to 1-2 cm. long near the apex of the cone. 
Seeds 6-8 cm. long, 4-5-3 cm. thick, outer coat bright orange-red when ripe. 

Distribution: Northern Territory: Central Australia in the Macdonnell and asso- 
ciated parallel range systems, sometimes in sclerophyllous communities with Eucalyptus 
or rarely Livistona, but sometimes as scattered plants on almost bare rocky slopes with 
Triodia species. 

Specimens examined: Noetheen Territoey: Alice Springs, R. H. PuUeine, vii.1917 
(NSW.40659); Standley Chasm, J. Garden and V. Lhuede, vii.1954 (NSW.40657), photo; 
Standley Chasm, N. Forde 864, 23.V.1957, and 10.iii.l957 (NSW.41470); Standley Chasm, 
L. A. S. Johnson, x.1957 (NSW., male cone); Central Australia, R. Tate, 1894 (NSW. 
40658). Also living material examined, Standley Chasm, L. A. S. Johnson, x.1957. 

M. macdonnellii is clearly a relict species, closely related to M. riedlei but isolated 
from it by arid country and with sufficient consistently different characteristics to be 
given specific rank. 

Bentham (1873), with only a fragment before him, could not distinguish it from 
M. fraseri (- M. riedlei), while Schuster (1932) placed it in the synonymy of his 
"typical" M. preissii (= M. riedlei excluding M. dyeri) without comment. The species is 
characterized by the glaucous fronds, the massive female cones (to almost 20 kgm. in 
weight) with most of the sporophyll-spines obsolete, and the huge seeds. 

Popularly, M. macdonnellii has been much confused with the very restricted relict 
palm species Livistona mariae, with which it is associated at Palm Valley, but the 
Macrozamia is a much more widely distributed plant in Central Australia, found at 
times in very forbidding habitats, though always in the Macdonnell Range complex. 
The extraordinarily large seeds of this species may have evolved as a selective adapta- 
tion to the uncertain rainfall of the interior. Their great food reserves must serve 
to give seedlings a good start in life. Glaucous bloom is likewise most strongly 
developed in this species, again suggesting adaptive significance. 

4. M. MiQUELii (F. Muell.) A.DC, Prodr., 16, pt. 2 (1868), 535 (in part, as to lectotype) ; 

Miq. in Arch. N4erl., 3 (1868), 248, and in Versl. Meded. Koninkl. Akad., ser. ii, 3 

(1869), 51 (in part). 

Basionym: Encephalar'tos miquelii F. Muell., Fragm. Phytogr. Austral., 3 (1862), 
38 (in part, as to lectotype). 

Typification: "Ad ostium fluminis Richmond River; C. Moore. Ad sinum Moreton 
Bay; W. Hill. Ad flumen Fitzroy River sub circulo capricornu; A. Thozet." Mueller's 
original concept was strangely mixed; Moore's specimen belonged to the species later 
described as M. faiocettii C. Moore (sect. Parazamia), whereas Hill's and Thozet's 
represented the present most dissimilar species. The description was derived in part 
from each of these species and the confused concept survived by copying in later 
publications by A. De CandoUe, Miquel and Mueller himself (see refs. in Fl. N.8.W., 
pt. 1, in press) both under the name of Macrozamia/ Enceplialartos miquelii and the 
misapplied name M. tridentata (Willd. ) Regel. It was never explicitly resolved, but 



BY L. A. S. JOHNSON. 97 

in time the name M. miquelii came to be applied (e.g. by Bentham, Fl. Austral., 6 
(1873), 253, and F. M. Bailey, Queensl. FL, 5 (1902), 1504) to the present species alone, 
whilst the other species has been generally known as M. fawcettii. Schuster's (1932) 
treatment is wholly confused and irrelevant in this connection. 

The type folder in herb. MEL contains three sheets, as follows: (1) Rockhampton, 
Thozet (with old male cone). (2) Moreton Bay, collector not indicated (this includes 
female cone fragments in a packet, probably a later addition; the cone at least is not 
part of the type series since Mueller had no female material). (3) Richmond River, 
C. Moore (with seeds in a packet labelled in Mueller's hand: "M. miquelii. This 
represents the genuine species according to locality"). Of these, 1 and 2 are the present 
species and 3 is M. faivcettii C. Moore. I have no doubt that 1 is the original Thozet 
syntype indicated by Mueller as "Ad flumen Fitzroy River . . ."; the frond of 2 may 
be the Hill syntype, but the cone is not; the frond of 3 may or may not be the Moore 
syntype (which from the description must have belonged to M. faivcettii) , but the seeds 
at least are a later collection. Mueller's annotation on 3 does not mean that he 
deliberately selected this element as lectotype; his concept always remained quite 
confused. Since Mueller's description applies in equal parts to both elements and the 
only definitely authentic syntype extant is Thozet's, which moreover includes male cone 
material mentioned in the original description, and since this specimen represents the 
species currently known as M. miquelii, I choose it (Rockhampton, Thozet MEL) as 
Lectotype. 

Synonymy, description and further discussion; see Johnson in Anderson, Flora of 
N.S.W., part 1 (in press). The synonymy and nomenclatural history, quite apart from 
the original confusion discussed above, are bewilderingly complex. 

Chromosome number: 2n = 18. 

Distribution: Queensland and New South Wales: Central and southern coast 
<Iistricts of Queensland from Rockhampton southwards, and far North Coast of N. S. 
Wales (Upper Richmond River), in sclerophyll forest on poor sandy or stony soils. 

Specimens examined: Queensland: Rockhampton, Thozet (MEL, Lectotype); 
Berserker Range, Rockhampton, J. L. Boorman (N.S.W., seeds only) ; Fraser Island, 
per Queensland Forest Service, xii.1922 (TECH); Mount Perry, J. L. Boorman, vi'i. 
1912 (NSW.40624, 40625); Moreton Bay [W. Hill?] (MEL); Queensland (N.S.W., cone 
only, old collection labelled M. douglasii, prob. from F. M. Bailey). New South Wales: 
In low flat ground between the Upper Richmond and Clarence Rivers, C. Moore, 1861 
(NSW.40610, HoLOTYPE of 31. cylindrica C. Moore). Also studied from living and 
preserved cultivated material. 

M. miquelii, as here defined, includes M. douglasii W. Hill ex F. M. Bail, from 
Fraser Island and M. mountperriensis F. M. Bail, from Mt. Perry, inland from Bunda- 
berg. These have been regarded, epsecially in Queensland (see Johnson in Anderson, 
I.e., for references), as distinct from M. miquelii, while M. cylindrica C. Moore has been 
regarded as distinct in New South Wales, though it is little known outside cultivation. 
However, I can find no real discontinuity nor any greater variation within this series 
of populations than exists in M. communis in New South Wales or M. riedlei in 
Western Australia. M. douglasii represents luxuriant plants from the deep sands and 
wet climate of Fraser Island, and M. mountperriensis the other extreme of small plants 
from the more continental climate of Mt. Perry, while M. miquelii and M. cylindrica 
(between which there seems to be no difference at all) represent the more average 
forms from scattered localities in the coast districts. The differences in stature and in 
number of reduced basal pinnae, however, must be to some extent genetically determined, 
since cultivated plants preserve them to some degree. The same applies to local 
ecotypes of M. communis and M. heteromera cultivated in Sydney. On the other hand 
there is a clear morphological as well as geographic discontinuity between the popula- 
tions of M. miquelii and M. communis, though they show parallel variation. This is 
discussed further by Johnson in Anderson, I.e. I know of no hybrids of M. miquelii. 
The species has been cultivated in Australia and overseas, and has a plethora of name.s 
in every conceivable rank (see Johnson, in press). 

G 



• 9-8 THE FAMILIES OF CYCADS AND THE ZAMIACEAE OF AL'STEALIA, 

5. M. COMMUNIS L. Johnson, sp. nov. 

Typiflcation: About three miles west of Mossy Point, near Mogo, New South Wales, 
J. W. Vickery, iv.l958 (NSW.43071), female. Holotype. 

Caudex plerumque subterraneus (per contractione radicium seppultus) sed in solo 
minime profundo usque ad 1-2 m. altus, 30-60 cm. diametro. Frondes in corona usque 
ad 50-100 sed saepe pauciores, 70-200 cm. longae, petiolo (basi lanata expansa 
exclusa) 12-40 cm. longo, rhachi non torta (stirpibus insanis exceptis) plus minusve 
applanata, ad pinnas inflmas 8-20 mm. lata, supra vel concava vel convexa aliquando 
carinata, sulcis duobus lateralibus angustis e basibus pinnarum decurrentibus instructa, 
infra subangulato-convexa. Pinnae 70-130, patentes, angulo acuto prorsum directae, 
plurimae arctae sed inflmae saepe distantes, quam illis M. miguelii crassiores 
rigidioresque, integrae, rectae, lineares, eae longissimae 16-35 cm. longae, 3-15 inflmarum 
gradatim reductae spiniformesque, 4-12 mm. latae, infra vix prominule nervis 7-13 
striatae, apicem pungentem versus sensim angustatae, basi pallida arete contractae et 
in axillis callosae rugosaeque (in sicco), supra saturate virides baud vel vix nitentes, 
pagina inferiore sola stomatibus instructa. Coni feminei 1-6, masculi 1-10, axillares 
inter frondibus, basi axis conigeri cataphyllis spiniformibus (ad 18 cm. longis) instructa. 
Coni masculi cylindracei (vetustiores aliquanto curvati) 20-45 cm. longi, 8-12 cm. 
diametro (5-8 cm. in stirpibus depauperatis), sporophyllis cuneatis vel obovato-cuneatis 
2-4 cm. longis, 1-5-2-5 cm. latis, spinis planis erectis 0-1-5 cm. longis (longissimis 2-5 
cm. versus apicem coni) basi 5-12 mm. latis. Coni feminei cylindracei, 20-45 cm. 
longi, 10-20 cm. diametro, in axibus 8-30 cm. longis 2-3 (-4?) cm. crassis gesti, sporo- 
phyllis 4-7 cm. longis (stipite 3-4-5 cm. incluso) parte terminali expansa glauca 
maturitate plerumque latitudine (3-8-5 cm.) quam altitudine (1-5-4-3 cm.) saltem duplo 
majore, spinis planis erectis, ad basin coni brevissimis (1-2-5 cm. longis) versus apicem 
longissimis (4-8 cm.), basi 8-20 mm. latis. Semina 3-4-5 cm. longa, 2-3 cm. crassa, 
integumenti parte exteriore carnosa miniata. Chromosomata somatica 2n = 18. 

Description (English), specimens and full discussion: See Johnson in Anderson. 
Flora of N.8.W., part 1 (in press). 

Distribution: New South Wales: Coast and ranges from Macleay River to Bega 
and westward to the head of the Goulburn River, in dry sclerophyll forest, on sandy 
or stony soils. 

This is the plant, familiar to every botanist who has worked in the Sydney 
district, which has been wrongly known in New South Wales as "M. spiralis"; it is not 
the species which has been called "M. spiralis" in Queensland (see below, under M. 
spiralis and M. lucida). Despite frequent references to it in the literature under this 
and other misapplied names (see Johnson, in Anderson, I.e.), it has never been 
described as a new species, hence the present description. The specific epithet reflects 
the abundance of this species in many parts of the Coast districts of New South 
Wales. M. communis shows a good deal of individual and some local variation in 
size of organs, but is clearly discontinuous in nature from related species of section 
Macrozamia. The Type is characteristic of the most luxuriant forms from the South 
Coast. Despite its occurrence in places near stands of certain species of section 
Parazamia, I have not yet found any evident hybridism either in the wild or in 
cultivated collections. Nevertheless, to judge by the behaviour of other species (e.g. 
M. moorei), hybrids are to be expected. M. communis is cultivated in gardens in 
Australia and overseas. Its well-known vernacular name is usually spelt "Burrawang", 
but today pronounced "Burrawong". For a discussion of the past nomenclatural 
confusion surrounding this species see Johnson in Anderson, I.e. 

6. M. DipLOMEKA (F. Muell.) L. Johnson, stat. nov. 

Basionym: Encephalartos spiralis (Salisb.) Lehm. var. diplomera F. Muell., Fragvi. 
Phytogr. Austral., 5 (1866), 172 (in part, as to lectotype, see below). 

Typiftcation: ". . . ab amico Carolo Moore in montibus Wambungle Mountains ad 
flumen Castlereaghii detectam." In point of fact these collections were made by 



BY L. A. S. JOHNSON. 99 

Moore's collector W. Carron, but Moore sent duplicates to Mueller. The type sheet is 
in herb. MEL labelled: Castlereagh River at the Wambungle [sic] mountains [New 
South Wales], [W.] Carron [before 1866]. This is a mixed collection: the sheet bears 
three pieces of fronds, of which two belong to the species here defined as M. diplomera 
(these I choose as Lectottpe) and one belongs to the species here defined as M. 
heteromera C. Moore (which was of similar mixed typification, see below). Mueller's 
description and epithet ("two-parted") clearly fit the former species better. The 
actual locality, of course, is the Warrumbungle Mountains, in the eastern sandstone 
foothills of which this species is found. 

Synonymy: See Johnson in Anderson, Flora of N.S.W., part 1 (in press). 

Since Mueller's original description in varietal rank is very brief and rather 
informal (though valid), I now provide a new Latin description (for English descrip- 
tion, see Johnson in Anderson, I.e.): 

Caudex plerumque subterraneus, 20(?)-40 cm. diametro. Frondes in corona usque 
ad 50 (?), sed saepe pauciores, 60-120 cm. longae, petiolo (basi lanata expansa 
exclusa) 10-20 cm. longo; rhachi non torta, plus minusve applanata, ad pinnas infimas 
8-12 mm. lata, supra concaviuscula vel convexa (proxime saepe plus minusve carinata), 
sulcis duobus lateralibus angustis e basibus pinnarum decurrentibus instructa, infra 
subangulato-convexa. Pinnae 70-120, valde patentes, angulo acuto prorsum directae, 
plurimae arctae sed infimae 2-4 cm. distantes, rigidiusculae, omnes paucis apicalibus 
exceptis angulo acutissimo in segmentis duobus vix divergentibus dichotome divisae 
(plerumque versus basin pinnae sed in pinnis subapicalibus versus medium, rare 
segmento uno pinnarum nonnullarum infimarum ipso diviso vel sub apice unidentato), 
eae longissimae 15-20 (-25?) cm. longae, nonnullae infimarum gradatim abbreviatae 
spiniformesque, 5-10 mm. latae (segmentis 2-5-5 mm. latis), infra 6-13 nervis (in 
segmentis 3-7 nervis) vix prominule striatae, (pinnae segmentave) ad apiccm 
pungentem sensim angustatae, basi pallida fiavescentive constricta et in axillis callosae 
rugosaeque (in sicco), sinu furcae pinnarum etiam saepe calloso rugosiusculoque, supra 
virides non nitentes (in sicco saepe flavescentes), paginis utrisque stomatibus instructis. 
Coni non certe noti, probabiliter illis minoribus M. communis similes, axe conigero 
femineo fide auctorum ferrugineo-tomentoso [cataphylla juniora in speciebus plurimls 
tomentosa sunt. L.J.]. 

Distribution: New South Wales: Southern part of North-west Slopes, around 
Coonabarabran and the eastern foothills of the Warrumbungle Mountains and east to 
the Mooki River, in dry sclerophyll forest on sandy or stony siliceous soils. 

Specimens and further discussion: See Johnson in Anderson, I.e. 

This species, of which cones are unfortunately unknown, is noteworthy for its 
divided pinnae. This character and the amphistomatic fronds clearly distinguish it 
from M. communis, the smaller inland forms of which it otherwise resembles. It is 
remarkable that its range corresponds in part with that of M. heteromera, which also 
has divided and amphistomatic pinnae, but which is as clearly a member of section 
Parazamia as M. diplomera is of section Macrozamia. Hitherto, most collectors and 
systematists have failed to distinguish M. diplomera from M. heteromera, usually 
including also a third species, M. stenomera (sect. Parazamia), which has divided but 
hypostomatic pinnae and is found to the north-east of this area. These facts of 
distribution suggest causal correlation of some kind: probably plants with divided 
pinnae possess some selective advantage, or at least are certainly not at a disadvantage, 
in the regional environment, but it is further possible that there has been introgressive 
gene-flow between populations belonging to the two rather diverse sections. A detailed 
study of these populations and their genetics should be of interest. In the field I have 
only once seen M. diplomera, in passing, and have been able to study M. heteromera 
only in areas where M. diplo-mera is absent. An apparent hybrid between the two was 
once grown in the Sydney Botanic Gardens from seed from the Coonabarabran district. 
The various distinctions between M. diplomera and M. heteromera are discussed by 
Johnson in Anderson, I.e. 



100 THE FAMILIES OF CYCADS AND THE ZAMIACEAE OF AUSTRALIA, 

ii. Section Paeazamia (Miq. ) Miq. 

In Arch. N4erl., 3 (1868), 250, and in Vers!. Meded. Koninkl. Akad., ii, 3 (1869), 
541; amplified hereunder. 

Basionyni: Encephalartos Lehm. sect. Parazamia Miq. in Yersl. Meded. Koninkl. 
Akad., 15 (1863), 374. 

Typification: Miquel (1863) included only Encephalartos pauli-guilielmi (W. Hill 
& P. Muell.) F. Muell. in his new section. The type species of Macrozamia sect. 
Parazamia is thus M. pauli-ffuilielmi W. Hill & P. Muell. The section as I conceive 
it includes the species correctly called M. spiralis (Salisb.) Miq. Hutchinson (1924) 
gives this binomial as type-species of Macrozamia, but, as argued above (p. 87), M. 
riedlei (Pisch. ex Gaudich.) C. A. Gardn. should be the generic type. The sectional 
name Parazamia may be used in the present circumscription only on this interpretation. 

Synonymy (infrageneric) : 

Macrozamia sect. Poly orientates Schuster (nom. illegit. ) subsect. Acutae Schuster, 
Pflanzenr., IV, i (1932), 87 (1 choose M. pauli-guilielmi W. Hill & P. Muell. as 
Lectotype. This permanently disposes of the name. Schuster nowhere mentions 
Miquel's earlier sectional names). 

Macrozamia sect. Polyorientales subsect. Curvatae Schuster, I.e. (I select M. 
fawcettii C. Moore as Lectotype). 

Pinnae usually more or less curved, spreading to erect (secund), nerves prominent 
on the lower surface especially when dry, apices entire to 2-7-toothed in adult fronds, 
bases slightly callous but never rugose, mucilage canals absent from the pinnae (present 
as always in the larger organs of the plant), the lowest pinnae not or only slightly, or 
one or two irregularly, reduced. 

Small plants with few fronds, caudex wholly subterranean or only its crown 
protruding. Rhachis sometimes straight, but more often moderately to very strongly 
spirally twisted or strongly recurved or incurved near the tip, from rounded to flattened 
or even markedly concave above, rounded or less often somewhat angular beneath 
when dry. Cones small (females 10-25 cm. long, 6-10 cm. diam. ; males 8-25 cm. 
long), sporophyll-spines short to medium (5 cm.), spreading to erect. 

This very natural section is confined to eastern Australia, and its species show 
evidence of reduction and some degree of neoteny or carrying-over of semi-juvenile 
characteristics (prominent nerves, unreduced basal pinnae, toothed pinna-tips, lack of 
mucilage-canals in pinnae, rounded petioles, small caudices) into the adult state. 
Accordingly it is probably of relatively recent origin, that is to say, as an evolutionary 
line its adult character-complex has been evolved more recently than those characters 
preserved in and characteristic of sect. Macrozamia. Furthermore, speciation appears 
to be still in progress in sect. Parazamia (cf. the races of M. pauli-guilielmi) and some 
of the species are very closely allied. On the other hand the plants seem on the whole 
less successful than in sect. Macrozamia; the populations are diffuse, rarely forming 
close stands, though in such species as M. hetero7nera and M. secunda the total popula- 
tion is by no means small. Since the plants are scattered, small and relatively 
inconspicuous, and produce few cones at irregular intervals, they are frequently over- 
looked by collectors and good cone-material is lacking for some species. Since the 
cones seem to be very similar throughout the section, this is not such a taxonomic 
handicap as it may appear. 

The species tend to show a replacement pattern (Text-figure 3), but this is not 
without overlap, though no two species of the section form mixed stands except 
marginally. Some marginal hybridism seems to occur between M. secunda and M. 
heteromera, M. heteromera and M. stenomera, M. heteromera and M. pauli-guilielmi ssp. 
plurinervia, and possibly between M. secunda and M. spiralis. In the case of the three 
subspecies of M. pauli-guilielmi there is a large-scale breakdown. Species of sect. 
Parazamia at times grow with or near species of sect. Macrozamia and hybrids may be 
found, as in the case of M. lucida and M. moorei at Dalmorton. Several apparent 



BY L. A. S. JOHNSON. 



101 



chance inter-sectional hybrids have come up in the cycad beds in the Sydney Botanicr 
Gardens. No chromosome numbers are recorded in this section. 

One cannot arrange the eight species of the section neatly in subgroups, but certain 
of them do form rather close pairsi or triplets; these are: (a) M. lucida-M. spiralis-M. 
secunda (the two end members differ markedly, but each is close to M. spiralis), 
(6) M. heteromera^M. stenomera, (c) M. pauli-guilielmi-M . faivcettii. Other cross- 
relationships are also evident, however, and M. platyrachis does not seem to fit in any 
one group better than another. 




Text-flg-. 3. — Distribution of 3Ia<yt'osamia section Parazamia: 7: M. platyrachis; 8: M. 
lucida; 9: M. spiralis; 10: M. secunda; 11: M. hetermnera; 12: M. stenomera; 13: M. pauli- 
guilielmi (a: ssp. pauU-ffvMielmi, b: ssp. plurinervia, c: ssp. flexuosa) ; 14: M. fawoettii. 



7. M. PLATYRACHIS* F. M. Ball, in Queensl. Agric. Journ., 3 (1898), 356, and in Queensl. 

Flora, 5 (1902), 1503 (as "platyrhachis") ; Schuster, Pflanzenr., IV, i (1932), 99 

("platyrhachis") . 

Typificatiofi : Range near Planet Downs, Queensland, H. C. Brock-HoUinshead, 
male, Holotype (a female cone was also sent). This is in herb. BRI. There is an 
IsoTYPE (vegetative) in NSW (40970). 

Caudex mostly subterranean, to 60 cm. diam. according to Bailey, probably often 
less. Fronds few (to ca. 12?) in the crown, to ca. 100 cm. long, petiole ca. 15-25 cm. 
long excluding the expanded base (which Bailey describes as with "no woolly sub- 
stance"); rhachis not or only moderately twisted (to ca. 180°?), 13-18 mm. broad at 
the lowest pinnae, flattened above with the two lateral furrows very shallow and 
obscure even when dry, convex and angular (at least when dry) beneath. Pinnae ca. 
50 (sometimes more?), somewhat spreading to spreading-erect, more or less forwardly 
directed and twisted at the base, the upper ones fairly crowded but the lowest 2-3 cm. 
apart, rather stiff and very coriaceous, entire, from somewhat recurved-falcate to fairly 

• The original spelling- platyrachis should, I think, be retained. In 1902 and subsequently 
Bailey used platyrhachis but did not indicate this as a deliberate correction of an unintentional 
error. Neither form is particularly good Greek ("-rrh-" would be best), but "-rachis" is 
acceptable latinized Greek. 



102 THE FAMILIES OF CYCADS AND THE ZAMIACEAE OF AUSTRALIA, 

Straight, broad-linear, the longest 30-40 cm. long and the lowest ones not reduced, 
12-20 mm. broad with ca. 15-18 rather prominent nerves beneath, somewhat tapered, 
but finally rather abruptly rounded to the mucronate apex, contracted to the somewhat 
paler, rather decurrent but not or scarcely callous base, green and somewhat glossy 
above, with stomata confined to the lower surface. Cones not examined; the following 
is fi'om Bailey's data: Cones stalked, 1-several (at least in the males) per plant 
(presumably axillary amongst the fronds). Male cones cylindrical (probably immature 
ones measured), 7-5-10 cm. long, ca. 2-5 cm. diam., sporophyll-spines erect, slender, to 
1 cm. long (prob. longer at times). Female cone more or less cylindrical, ca. 16 cm. 
long (ca. 8 cm. diam.?) ; sporophyll-spines more or less erect, flattened, the longest 
(towards the apex of the cone) ca. 1-3 cm. long. Seeds ca. 2-5 cm. long, outer coat 
reddish ("reddish-brown" ace. to Bailey, these probably not fresh). [Doubtless cones 
and their parts vary a good deal in size, as in other species.] 

Distribution: Queensland: Dryish (not arid) parts of Central Queensland (Planet 
Downs) in hilly country, doubtless in dry sclerophyll communities. 

Specimens: See above, under Typiflcation. 

M. platyrachis is a poorly known but quite distinctive species. It is the most 
northerly member of section Parazamia and is readily recognized by its very broad, 
stiffly coriaceous pinnae and the broad rhachis which is flattened above and angular 
beneath. It does not show particular aflinity with any one species, though clearly 
belonging to the section. Schuster (1932) grouped it with M. fawcettii because of its 
broad pinnae, but this is surely a case of convergence or parallel evolution in a single 
character; in other respects it does not much resemble M. fawcettii. 

Bailey's statement that the petiole-bases lack woolly hairs needs conflrmation; 
so does the collector's remark that the species does not cause "rickets" in stock. 

8. M. LuciDA L. Johnson, sp. nov. 

Typiflcation: Southern side of Ngungun, abt. 400 ft. alt., Glass House Mountains, 
Queensland, L. A. S. Johnson, 13.vi.l951 (NSW.40668), vegetative. Holotype. (Since 
no good material with conea was available, I have nominated as the type this specimen 
which I have seen living in the field.) 

Caudex plerumque subterraneus, 10-20 (-30?) cm. diametro. Frondes in corona 
plerumque 2-15 (nonnunquam usque ad 40?), 80-110 cm. longae, petiolo (basi lanata 
expansa exclusa) 25-50 cm. longo; rhachis non torta, teretiuscula, ad pinnas infimas 
3-7 mm. lata, supra saepissime rotundato-convexa (sed nonnunquam applanata vel 
proximo plus minusve medio subsulcata), sulcis duobus lateralibus angustis e basibus 
pinnarum decurrentibus instructa (in sicco), infra semper rotundato-convexa (baud 
angulata). Pinnae 50-100, patentes (seriebus duabus tamen non in eodem piano) et 
presertim versus apices suos decurvatae, plurimae angulo acuto prorsum directae sed 
infimae cum rhachi angulum rectum saepe formantes, basi tortae, plurimae arctae sed 
infimae saepe 1-5-3 cm. distantes, integrae, pleraeque plus minusve recurvato-falcatae 
(sed versus apices suos aliquando prorsum curvatae), lineares, eae longissimae 15-35 
cm. longae, infimae baud vel vix abbreviatae, 7-11 mm. latae, infra nervis prominentibus 
5-11 striatae, apicem pungentem versus sensim angustatae, basi conspicue pallida 
callosiuscula sed non rugosa constrictae, supra nitentes virides, pagina inferiore sola 
stomatibus instructa. Coni maturi non visi, probabiliter illis M. spiralis similes; 
sporophyllis masculis usque ad 4 cm. longis, spinis brevibus. Conos ad M. lucidam 
probabiliter pertinentes, F. M. Bailey (1902) sequens nunc describo: Axis coniger usque 
ad 30 cm. longus. Coni masculi cylindracei, 15 cm. longi (vel longiores), ca. 4 cm. 
diametro, [immaturi? L.J.] spinis sporophyllorum infimorum obsolescentibus eis 
sporophyllorum apicalium usque ad 1-2 cm. longis. Coni feminei 15-20 cm. longi, 
7-5-9 cm. diametro, sporophyllorum parte terminali ad 4 cm. lata et 2 cm. alta, spinis 
0-6-5 cm. longis longissimis versus apicem coni dispositis. Semina ca. 2-5 cm. longa, 
ca. 2 cm. crassa, integumenti parte exteriore rubella. 

Description (English) and discussion of misapplied names: See Johnson in 
Anderson, Flora of N.B.W., part 1 (in press). 



BY L. A. S. JOHNSON. 103 

Distribution: Queensland and New South Wales: Southern Coast region of Queens- 
land from Wide Bay to Moreton District, and Clarence River (N. Coast of N.S.W.). 

Specimens examined: Queensland: Southern side of Ngungun, abt. 400 ft. alt.. 
Glass House Mountains, L. A. S. Johnson, 13.vi.l951 (NSW.40668) ; Mt. Coonowrin, 
Glass House Mountains, ca. 300 ft, C. E. Hubbard 4112, 21.ix.l930 (BRI); Taylor's 
Range, near Brisbane, C. T. White, i.l912 (BRI); Palmwoods, C. T. White, 6.V.1907 
(BRI); Enoggera Creek, F. M. Bailey, 7.i.l875 (BRI); Enoggera, C. T. White, vi.l919 
(NSW.40699); Cedar Creek, near Brisbane, C. T. White 1961, 6.V.1923 (NSW.40671); 
Brisbane, J. L. Boorman, iv.l899 (NSW.40670). New South Wales: Three miles E. 
of Dalmorton, L. A. S. Johnson and E. P. Constable, vi.l957 (NSW.43069). Living and 
preserved cultivated material also examined. 

This is the species which has been wrongly known in Queensland, but not in 
New South Wales, as M. spiralis (see below, under M. spiralis). Though related to the 
true M. spiralis, it clearly constitutes a quite distinct series of populations, and is 
readily distinguished in cultivation as well as in the wild. M. lucida is characterized 
by its long slender petioles, not or scarcely twisted rhachis, and the curved, very glossy 
(whence the specific epithet) pinnae with sharply demarcated whitish but not very 
callous bases. Though it does not form dense stands it is widely spread in hilly 
country of the coast districts of southern Queensland, but in New South Wales only 
a single small stand is known as yet. It has been cultivated in the Sydney Botanic 
Gardens for many years and there maintains its distinctive appearance. For further 
treatment see the forthcoming Flora of New South Wales, part 1, where a probable 
hybrid population derived from M. lucida and the very different M. moorei is also 
discussed. 

9. M. SPIRALIS (Salisb.) Miq., Monogr. Gycad. (1842), 36, as to basionym, excl. descr. 

and fig.; misapplied also by all later authors. 

Basionym: Zamia spiralis Salisb., Prodr. Stirp. (1796), 401. 

Typification: "Sponte nascentem juxta Port Jackson, locis umbrosis, legit Dav. 
Burton." As with most of Salisbury's types, no specimen of this appears to be extant. 
The correct application of the name must be established, if possible, from the descrip- 
tion and other evidence. The description reads: "Z. foliolis 30-40 jugis, extrorsum 
falciformibus apice spinose 3-5 dentatis." "Petioli paululum spirales sunt." Together 
with the information on locality, habitat and collector quoted above, which was inter- 
polated by Salisbury between the first and second descriptive sentences, this is the 
whole of the original information. The first author to do more than repeat Salisbury's 
description was R. Brown, Prodr. (1810), 348, who applied the name Z. spiralis to a 
mixture of the species now treated as M. communis L. Johnson and M. riedlei (Fisch. 
ex Gaudich. ) C. A. Gardn., but expressed doubt whether these were conspecific. Miquel 
(1842), in making the new combination under Macrozamia, described only plants of 
M. riedlei (from figures of F. Bauer), but quoted R. Brown for the Port Jackson 
locality (representing M. communis) , and, of course, his combination is nomenclaturally 
based on Salisbury's name, of which he had not seen the type. However, he remarked 
that Brown's description and Bauer's figures did not agree with Salisbury's description 
of dentate-tipped leafiets, and added "Quam ob rem credere posses, speciem ab his 
auctoribus recensitam a vera Zamia spirali Salisl). differre", with further remarks 
illustrating this doubt and also the confusion with the garden plants known as 
Enceplialartos tridentatus and E. pungens (these were based, as Zamia tridentata Willd. 
and Z. pungens Ait., on young cultivated plants of supposed South African origin and 
of quite uncertain identity. Juvenile plants of Macrozamia and Enceplialartos cannot 
be determined to the species). 

Later, in Yersl. Meded. Koninkl. Akad., 15 (1863), 368-370, Miquel made it clear 
that Bauer's plates (from which his earlier description of M. spiralis had been taken) 
in fact depicted his Enceplialartos fraseri (- Macrozamia riedlei) and went on to treat 
as "Enceplialartos spiralis [Salisb.] Lehm." a mixture of M. communis and M. miquelii. 



104 THE FAMILIES OF CYCADS AND THE ZAMIACEAE OF AUSTEALIA, 

In a similar treatment, but under Macrozamia (Mlquel, 1868, p. 249; 1869, p. 52), he 
again expressed doubt regarding Salisbury's plant. 

Other authors have applied the name M./E. spiralis in various ways, but most 
generally to M. communis (e.g. Mueller, passim, Bentham, and such N. S. Wales authors 
as C. Moore, Maiden and Betche, Brough and Taylor. Full refs. in Fl. N.8.W., Part 1, 
in press), though they often included one or more other species in their concept. 
Queensland botanists, following F. M. Bailey (esp. in Queensl. Fl., 5 (1902), 1504), have 
used M. spiralis for the quite different species M. lucida L. Johnson. 

Schuster, Pfianzenr., IV, i (1932), 88, ignored priority and synonymized M. spiralis* 
and Z. spiralis* under his extraordinarily inclusive and confused concept of "M. 
tridentata (Willd.) Regel", based, of course, on the unidentifiable Z. tridentata Willd. 
(1806). 

No one apart from Miquel (1842, 1869) seriously considered the original application 
of Salisbury's name. This we must now do. 

Two species grow sufficiently close to Port Jackson to have been collected by 
David Burton before 1796. These are M. communis L. Johnson ("ikf. spiralis" of N.S.W. 
botanists) and the species hitherto known as M. corallipes Hook. f. While it would 
be convenient to retain the usage traditional in N. S. Wales (though not in Queensland), 
an honest application of the International Code of Botanical Nomenclature will not 
allow this.f In the absence of a material type, a name's application must be determined 
from the author's words and associated facts. Only when a reasonable degree of 
certainty is impossible does the Code permit a name to be dropped. 

Juvenile fronds of all species have toothed pinnae, but Salisbury's description 
cannot apply to a wholly juvenile plant, since he mentions 30-40 pairs of pinnae. As in 
section Macrozamia generally, plants of M. communis old enough to have 60-80 pinnae 
in their fronds never have falciform pinnae spinosely 3-5-toothed at the apex, nor is 
the petiole (or rhachis) spiral (even "paululum"). On the other hand this description 
applies perfectly to many submature individuals of M. corallipes (sect. Parazamia). 
Consequently I have no doubt that Salisbury's plant was the latter species, for which 
the name Macrozamia spiralis (Salisb.) Miq., being nomenclaturally based on Zamia 
spiralis Salisb., must henceforth be used. The implications of this in the typification 
of the genus Macrozamia Miq. are discussed above (p. 87). 

Synonymy, description and specimens: See Johnson in Anderson, Flora of N.S.W., 
part 1 (in press). The synonymy and history of misapplication are complex. 

Distribution: New South Wales: North-east part of Central-western Slopes to 
Central Coast, from Dunedoo and the Goulburn River Valley to the lower Blue 
Mountains, Sydney and Waterfall districts, in dry or rarely semi-wet sclerophyll forest 
on poor, sandy or gravelly soils. 

M. spiralis, as here defined, includes only those populations which would have 
been referred in the past to M. corallipes Hook. f. The plants are usually scattered and 
inconspicuous, with few fronds. The semi-juvenile fronds often seen, especially after 
fires, may be spirally twisted through 180° or more, and often have relatively few, 
widely-spaced pinnae with 2-several small teeth near their tips. Fronds of fully 
adult form, however, are little twisted and have more crowded pinnae usually simply 
mucronate at the tips. Occasionally M. spiralis grows close to stands of M. communis, 
but no hybrids are known. It overlaps M. pauli-guilielmi ssp. plurinervia in the Goulburn 

* He cited both of these as "ex parte", which is meaningless for Z. spiralis Salisb., 
described from a single specimen. He nowhere mdicated what he imagined the other part 
of this to be. 

t Every responsible taxonomist, to say nothing of other botanists, must be disturbed by 
the number of such irritating corrections of ancient mistakes which the present rules of 
nomenclature make necessary, but the remedy cannot lie in individual botanists' refusal to 
obey the Code. In any case the nomenclature of Macrozamia has been so confused that a 
new point of departure should be welcome. I hope the present revision will satisfactorily 
supply this. 



BY L. A. S. JOHNSON. 105 

River Valley, but I have found no actual contact of populations. M. spiralis is replaced 
to the west by the closely allied M. secunda, and one or two inadequate specimens 
suggest that some interbreeding may occur, though I have seen no living intergrading 
plants. 

It is important to remember that morphological studies (e.g. by Brough and Taylor, 
1940) and chromosome counts under the name of M. spiralis apply in fact to 31. 
communis. 

10. M. SECUNDA C. Moore in Journ. Roy. Soc. N.S.W., 17 (1884), 120; Moore & Betche, 

Handb. Fl. N.S.W. (1893), 379, excl. var.? 

Typiflcation:* "Near Reedy Creek, east of Mudgee, where it was first found in 
1858, but without fruit. Again found with only one old fruiting cone not far from 
Dubbo, by Rev. J. Milne Curran, in 1883. . . ." Only the former collection is extant, in 
herb. MEL, labelled: Reedy Creek, C. Moore, 1858. I therefore select it as Lectotype. 

Synonymy, description, specimens and discussion: See Johnson in Anderson, Flora 
of N.S.W. , part 1 (in press). 

Distribution: New South Wales: Central-western Slopes and lower parts of Central 
Tablelands, from near Gilgandra to Grenfell and east to the Main Divide from Mudgee 
to Capertee, in dry sclerophyll forest on sandy or stony country. 

M. secunda and M. spiralis may be regarded as a pair of vicarious species. The 
distinction seems sharp enough to justify specific rank. The concave petiole and stiff, 
usually narrow, secund pinnae with very crowded nerves are characteristic. As in 
other species found in dry localities, the pinnae have much sclerenchymatous tissue 
and some glaucous bloom. Some hybridism with M. heteromera is evident where the 
species meet, as discussed by Johnson in Anderson, I.e. The name M. secunda var. 
dichotoma C. Moore & Betche, Handb. Fl. N.S.W. (1893), 379, probably applies to such 
a hybrid, but the type has been lost. 

11. M. HETEROMERA C. Moore in Journ. Roy. Soc. N.S.W., 17 (1884), 122 (in part, as 
to lectotype and excl. var. tenuifolia) ; Moore & Betche, Handb. Fl. N.S.W. (1893), 
380, in part. 

Typiflcation: "Among the Warrenbungle ranges and on the Castlereagh River 
country. Discovered in 1858; since collected near Rocky Glen, between Coonabarabran 
and Gunnedah." An old sheet dating from Moore's time, labelled "Warrumbungle 
Ranges", without collector's name, but probably collected by W. Carron or by Moore, 
is in NSW (40720). No specimen from Rocky Glen is extant. No original cone 
material has been preserved. I consider that NSW. 40720 is the type sheet, but it bears 
a mixture, consisting of two pieces of frond belonging respectively to M. heterom-era 
as here defined, and M. diplomera (F. Muell.) L. Johnson. Moore's description and 
comments covered both species, but specimens and cultivated plants later named by 
him chiefly belong to the former. Consequently I select as Lectotype that part of 
NSW. 40720 representing M. heteromera as herein defined. Both M. diplomera (see 
above, page 98) and M. heteromera are thus based on a mixture of the same two species, 
probably from duplicates from the same original collection, but selection of lectotypes 
in accordance with the emphasis of original descriptions or later usage allows both 
names to be used. The respective epithets ("two-parted", and "different-" or "variably- 
parted") are particularly appropriate in their present application. Moore's total concept 
was extended by two varieties (see below). 

Synonymy: 

M. heteromera var. glauca C. Moore in Jonirn. Roy. Soc. N.S.W., 17 (1884), 122. 

M. spiralis (Salisb.) Miq. var. heteromera (C. Moore) Maid. & Betche, Census 
(1916), 9, in part. 

* Moore made no reference to M. sijiralis (SalLsb.) Miq. var. ? secunda Benth., Fl. Austral., 
6 (1873), 252, which is therefore not the basionym but an earlier synonym in a different rank. 
Maiden and Betche, Census (1916) 9, later made the homonymous varietal combination based 
on M. secunda C. Moore. 



106 THE FAMILIES OF CTCADS AND THE ZAMIACEAE OF AUSTRALIA, 

M. heteroviera var. tenuifolia Schuster* forma harmsii Schuster, Pflanzenr., IV, 
i (1932), 96. 

Description, specimens and further discussion: See Johnson in Anderson, Flora of 
N.S.W., part 1, in press. A specimen from Warialda, per Glenfleld Veterinary Research 
Station, ix.l958 (NSW.46090) was collected too late for citation in the Flora. 

Distribution: New South Wales: North-western Slopes, in the eastern and southern 
Pilliga Scrub from Narrabri to Coonabarabran and the foothills of the Warrumbungle 
Mountains, with an apparent outlier to the north-east near Warialda and Howell, in 
dry sclerophyll forest on siliceous soils. 

M. heteromera resembles M. stenomera in its dichotomously divided pinnae, but 
these differ in the consistent presence of stomata on the upper surface. The degree of 
division varies considerably, both individually and locally, but is on the average less 
than that in M. stenomera. In cultivation, Moore's var. glauca preserves its characters 
of rather broad, once-divided, glaucous pinnae, but plants can be found showing every 
gradation in these characters and the glauca forms do not seem to show sufficient 
geographic or ecological cohesion to be treated as a subspecies. Var. tenuifolia C. Moore 
is M. stenomera (q.v.). The distinction from and possible hybridism with M. diplomera 
are dealt with above (p. 99). Some apparent hybrids with the adjacent M. secunda 
and M. stenomera are discussed by Johnson in Anderson {I.e.). Probable hybrids with 
M. pauli-guilielmi ssp. plurinervia also occur, near Warialda. 

Islote added 11th June, 1959. — A recent collection, from 19 miles east of 
Coonabarabran on Gunnedah road, W. Morris, v. 1959 (NSW.47164), has the pinnae 
quite undivided, and in general appearance resembles M. spii-alis, but has the 
amphistomatic pinnae and short petiole (11 cm.) of M. heteromera. The locality is 
within the range of the latter species and about fifty miles north of the nearest known 
M. spiralis. The intervening areas may possibly reveal some intergradation between 
the two species. 

12. M. STENOMERA L. Johnson, sp. nov. 

Typification: Above Coryah Gap at 4500 ft., Nandewar Mountains, New South 
Wales, L. A. S. Johnson and E. F. Constable, xi.l954 (NSW.32204), vegetative. 

HOLOTTPE. 

Synonymy :f 

M. heteromera C. Moore var. tenuifolia C. Moore in Journ. Roy. 8oc. N.S.W., 17 
(1884), 122 (not M. tenuifolia Hort. ex Miq. in Versl. Meded. Koninkl. Akad., ser. ii, 3 
(1869), 55, in synon., nom. invalidum). 

M. heteromera var. tenuifolia Schuster, Pflanzenr., IV, i (1932), 96, nom. illegit, 
in part, excl. forma harmsii Schuster (this var. is described as new, not taken from 
C. Moore). 

M. heteromera var. dicranophylloides Schuster, I.e. 

Caudex plerumque subterraneus, ca. 8-15 cm. diametro. Frondes in corona paucae 
(ca. 2-10), 40-80 cm. longae, petiolo (basi lanata expansa exclusa) 7-15 cm. longo; 
rhachi plerumque praesertim apicem versus plus minusve torta (per 90°-360°) etiam 
plus minusve recurvata vel sinuosa, ad pinnas inflmas (4-) 5-9 mm. lata, supra proximo 
plana vel concaviuscula vel nonnunquam aliquantulum carinata sed versus apicem 
rotundato-convexa, sulcis duobus lateralibus angustis e basibus pinnarum decurrentibus 
instructa (in sicco), infra plerumque rotundato-convexa. Pinnae 70-120, patentes 
(seriebus duabus tamen non in eodem piano) vel suberectae, plus minusve prorsum 
directae, basi tortae, plurimae arctae sed infimae 1-3-4 cm. distantes, quam illis M. 

* See under M. stenomera^ below. 

t Since the synonyms are all in varietal rank and of confused past application, and 
since they were based on a specimen rather than a population-concept, I prefer (as the 
I.e. B.N. permits) to use a new epithet and a new type \n describing the taxon in specific 
rank. This epithet may be legitimately retained in the rank of subspecies, should further 
knowledge show this to be more appropriate. 



BY L. A. S. JOHNSON. 107 

heteromerae laxiores, omnes (apicalibus aliquando exceptis) plerumque versus bases 
suas in segmentis angusto-lineai'ibus plus minusve divergentibus 1-4-plo (pleraeque 2-3- 
plo) dichotome divisae, longissimae 10-20 cm. longae, inflmae baud vel vix abbreviatae 
nunquam spiniformes; segmentis primariis 2-4 mm. latis (secundariis tertiariisque 1-3 
mm.) infra nervis prominentibus 4-6 (2-5 in segmentis secundariis tertiariisque) 
striatis, (segmentis ultimis) in apicem mucronatum vel aliquando bidentatum 
augustatis vel ultime abruptiuscule rotundatis, (pinnae) basi flavescenti aurantiacave 
callosiuscula sed non rugosa constrictae (callo in sinu furcarum pinnarum vel obsolete 
vel parvo), supra virides vix nitentes (sed quam illis M. heteromerae viridiores), infra 
non glaucae, pagina inferiore sola stomatibus instructa. Coni non certe noti, probabiliter 
illis M. heteromerae specierumque aliarum sectionis Parazamiae similes. 

Description (English), specimens and further discussion: See Johnson in Anderson, 
Flora of N.S.W., part 1 (in press). 

Distribution: New South Wales: North-western Slopes and Northern Tablelands 
outlier in and around the Nandewar Mountains, in dry sclerophyll forest, on stony but 
not highly siliceous soils. 

In the past these plants have not been distinguished from M. heteromera, but 
they appear to constitute a reasonably distinct group of populations which I have 
accordingly treated as a distinct species. Definitely matched cones are unfortunately 
lacking, but within section Parazamia these organs differ little from species to species 
and, though their collection is much needed, they probably have little bearing on the 
independent status of M. stenomera. The specific epithet refers to the characteristically 
narrow segments of the pinnae which, together with their usually higher degree of 
division, lack of glaucousness and especially lack of stomata on the upper surface, 
distinguish the species from M. heteromera. Spiral twisting of the fronds is also 
generally more pronounced in M. stenomera. The species grows on less siliceous soils 
than M. heteromera. 

Specimens of apparent hybrids of M. stenomera with the adjacent M. heteromera 
are discussed by Johnson in Anderson, I.e. 

13. M. PAULi-GuiLiELMi* W. Hill & F. Muell. in F. Muell., Fragm. Phytogr. Austral., 

1 (1859), 86. 

Typification: "In vicinia sinus Moreton Bay rara. W. Hill" (MEL, not seen but 
unmistakable from the description). 

As here defined this species comprises three geographic races, the extreme forms 
of which differ from each other quite as much as do several other population-groups 
treated in this revision as full species. However, in the latter cases the populations 
are effectively discontinuous (though some limited local interbreeding may occur), 
whereas the contiguous races of M. pauli-guiliehni show no such discontinuity but 
rather a clinal transition. Each of the three races, however, has decided characteristics 
of its own (preserved in cultivation under identical conditions) and occupies an area 
of distinct ecological character. Furthermore, the geographically intermediate race is 
by no means simply intermediate morphologically between the two terminal races, but 
is more extreme in certain features than either of them. We may reasonably infer 
either (a) that three populations have been effectively isolated in the past (later 
Tertiary and/or Pleistocene?), during which time they became genotypically and 
phenotypically differentiated, but that subsequent breakdown of isolation has led to 
extensive gene interchange in broad transitional contact zones, or (b) that differentia- 
tion has taken place under strong (ecological) selection pressure within a widespread, 
at least originally more or less panmictic population of high potential genetic 

* rjnlike such epithets as "fawcetti", here corrected according- to the I.C.B.N. to "fawcettii", 
the epithet pauli-guiUehni is a direct Latin g-enitive (from "Paulus Guillelmus", i.e. Paul 
Wilhelm, Prince of Wurtemburg-) and the terminal "i" should not be doubled. The clumsy 
epithet has been spelt in several ways (see Fl. N.S.W., pt. 1) but the original orthography 
"Pauli Guilielmi" is not to be altered except by decapitalization and hyphenation, as 
authorized by the I.C.B.N. 



1,08 THE FAMILIES OF CYCADS AND THE ZAMIACEAE OF AUSTRALIA, 

variability, but without development of a fully effective breeding-barrier (genetic or 
spatial) between the population segments. (There is no evidence of effective intro- 
gression from any other species, another theoretically possible cause of intergrading 
geographic races.) In either case the three segments at the present time cannot be 
regarded as specifically distinct in nature, but each of them has sufficient cohesion and 
ecological and morphological distinctness to be treated as a subspecies.* 

Full descriptions, synonymy and discussions of M. pauli-guilielmi and its three 
subspecies, and citation of New South Wales collections, will be found in the forth- 
coming Flora of N.S.W., part 1. Only the following key, enumei'ation and necessary 
Latin diagnosis are given here. 

1. Petiole much flattened, 5-15 cm. long. Pinnae lax, numerous (140-200), pale at base. 
Broadest pinnae 2-4 (-5) mm. broad, 3-5-nerved (odd ones rarely to 7). S. Queensland 
a. ssp. pauli-guilielmi. 

1.* Petiole flattened to rounded, 5-30 cm. long, if flattened and less than 15 cm. long then 

broadest pinnae 4-7 mm. broad and 6-10-nerved and usually pinkish at the base. 

2. Petiole 5-20 (-25) cm. long, 5-11 mm. broad at lowest pinnae. Pinnae concave to 

convex, the broadest 4-7 mm. broad (10-30 cm. long), 6-10-nerved, often rather stiff, 

pinkish at the base when living. Extreme S. Queensland and N. Tablelands, N.W. 

Slopes and Upper Hunter Valley (N.S.W. ) &. ssp. plurinervia. 

2.* Petiole 20-30 cm. long, 4-8 mm. broad at lowest pinnae. Pinnae concave, the broadest 
3-5 mm. broad (17-30 cm. long), 5-6 (-7) -nerved, lax, pale at the base. Manning River- 
Lake Macquarie (N.S.W. ) c. ssp. flexiiosa. 

a. ssp. PAULI-GUILIELMI. 

Typification: As for the species. 

Synonymy, description and further discussion: See Johnson in Anderson, Flora of 
N.S.W., part 1 (in press). 

Distribution: Queensland: Southern areas, somewhat away from the coast, in 
the Districts of Burnett, Darling Downs and western portion of Moreton. 

In the southernmost part of this range the subspecies shows clinal intergradation 
with ssp. plurinervia, and such intergrades extend into the far north of New South 
Wales, for instance near Acacia Creek. Plants of this subspecies have been cultivated 
a good deal, both in Australia and abroad. 

b. ssp. PLUEiNEKviA L. Johnsou, subsp. nov. 

Typification: "Reedy Creek" Station, near Bonshav/, New South Wales, J. Leader, 
iv.l956 (NSW. 40958), male. Holotype. (Female cone pieces are also associated with 
this sheet.) 

Frondes (30-) 40-90 cm. longae, petiolo (5-20, rare 25 cm. longo, basi expansa 
exclusa) cum rhachi supra applanato vel plus minusve rotundato vel aliquanto sulcato, 
infra angulato rotundatove, ad pinnas infimas 5-11 mm. lato, rhachi valde torta per 
revolutionibus una vel pluribus (vel in frondibus perbrevibus plantarum depauperatarum 
per ca. 180°). Pinnae 50-150, arctae vel sparsiusculae, 10-30 cm. longae, rigidae vel 
laxiusculae, 4-7 mm. latae, 6-10-nerviae, aliquando plus minusve glaucae. basibus 
plerumque aurantiacis rubellisve, supra concavae vel planae etiam convexiusculae. Coni 
plerumque glauci, feminei saepissime ovoidei spinis plerisque sporophyllorum patento- 
erectis. 

Description (English), specimens and further discussion: See Johnson in Anderson, 
Flora of N.S.W., part 1 (in press). 

* The category of variety, which many taxonomists still use to cover variations of 
diverse nature and of very unequal and frequently undefined biological and' evolutionary 
significance (often not populations at all), usually conveys no more than that an author 
has chosen to name individuals differing in certain ways from the nomenclatural type, itself 
an object of no biological significance. Subspecies, on the contrary, though used by few 
taxonomic botanists in this country, has come in recent years to have a fairly definite 
meaning, at least to those whose interest is in evolutionary processes. This concept of an 
ecogeographic segment of a species has nothing in common with Schuster's (1932) virtually 
meaningless "subspecies" in Macrogamia. 



BT L. A. S. JOHNSON. 109 

Distribution: Queensland and New South Wales: Extreme south of Darling Downs 
District (Queensland) and lower parts of N. Tablelands, North-western Slopes and 
Upper Hunter Valley (New South Wales), in dry sclerophyll forest or woodland on 
stony slopes. 

Intergrading forms to ssp. pauli-guilielmi are found in the extreme north of this 
region (see above). Hunter Valley populations show a clinal approach to ssp. flexuosa. 
Plants of the ssp. plurinervia series of populations have usually been referred to M. 
flexuosa or M. pauli-guilielmi. A full discussion is given in Flora of N.8.W., part 1. 

c. ssp. FLEXUOSA (C. Moorc) L. Johnson, comb, et stat. nov. 

Basionym: Macrozamia flexuosa C. Moore in Journ. Roy. Soc. N.8.W., 17 (1884), 
121. 

Typiflcation: Moore says only ". . . grows plentifully between Raymond Terrace 
and Stroud". There is only the following collection from Moore's time and this district 
in herb. NSW, and I consider it to be the Holotype: Limeburner's Creek, New South 
Wales, E. Betche, i.l883 (NSW.40951), vegetative, associated cones apparently lost. 

Synonymy, specimens and further discussioyi: See Johnson in Anderson, Flora of 
N.S.W., part 1 (in press). 

Distribution: New South Wales: Southern part of North Coast and extreme north 
of Central Coast, from Bulladelah district to Lake Macquarie, in sclerophyll forest, on 
rather siliceous soils. 

In some respects this race has more resemblance to the geographically remote ssp. 
pauli-guilielmi than to the contiguous ssp. plurinervia, although it differs markedly 
from ssp. imuli-guilielmi in the long, rounded petioles. As stated above, the southern 
populations referred to ssp. plurinervia exhibit clinal approach to ssp. flexuosa. 

Submature plants of M. spiralis with rather twisted fronds have often been 
referred to M. flexuosa, but true M. pauli-guilielmi ssp. flexuosa is distinguished by its 
fronds being twisted through at least two complete revolutions and by its concave 
pinnae. No area of contact with M. spiralis is known, but the two may well meet in 
the Wollombi Creek-Macdonald River district. 

14. M. FAWCETTii C. Moore in Journ. Roy. Soc. N.8.W., 17 (1884), 120. 

Typiflcation: "On high ground on the upper part of the Richmond; discovered by 
C. Fawcett, Esq., P. [olice] M.[agistrate], . . . Only freshly-gathered leaves and old 
male cones have as yet been seen." This is in herb. NSW, labelled: Upper Richmond 
River, New South Wales, C. Fawcett (NSW.40653), vegetative with male cone fragments. 
Holotype. 

Synonymy, description, specimens and discussion: See Johnson in Anderson, Flora 
of N.S.W., part 1 (in press). 

Distribution: New South Wales: Farther North Coast and ranges, from Richmond 
River to Coff's Harbour, in dry or semi-wet sclerophyll forest on siliceous soils. 

M. fawcettii, though closely related to M. pauli-guilielmi, the other species with 
twisted fronds, is isolated geographically and is clearly distinguished from all forms 
of M. pauli-guilielmi by its very broad, usually several-toothed pinnae. Whatever might 
happen if it were brought into contact with M. pauli-guilielmi, it is at the present time 
effectively isolated as a species. It overlaps M. lucida in distribution, but no contact 
has been observed. 

The nomenclatural confusions which ensued from Mueller's association of specimens 
of M. fawcettii with the very different M. miquelii are discussed under the latter species 
and in the forthcoming Flora of N.S.W., part 1. 

III. BowENiA Hook, ex Hook. f. 
Hook, ex Hook, f.* in Bot. Mag., 89 (1863), sub t.5398; ibid., 98 (1872), sub t.6008; 
F. MuelL, Fragm. Phytogr. Austral., 5 (1866), 171; A.DC, Prodr., 16, pt. 2 (1868), 

* J. D. Hooker described the genus and species, but adopted names proposed though nut 

published by W. J. Hooker, whom he again cited as author in 1872. Later authors, confused 

by this, have cited either "Hook." or "Hook. f.". Under the I. C.B.N, the correct citation is 
as given above. 



110 ■ THE FAMILIES OF CYCADS AND THE ZAMIACEAE OF AUSTRALIA, 

524; Miq. in Arch. N6erl., 3 (1868), 254, and in Versl. Meded. Koninkl. Akad., ser. ii, 
3 (1869), 57; Benth., Fl. Austral., 6 (1873), 254; Regel in Acta Hort. Petrop., 4 (1876), 
316; F. M. Bail., Queensl. Fl., 5 (1902), 1507; Chamberlain in Bot. Gaz., 54 (1912), 419; 
Schuster, Pflanzenr., IV, i (1932), 85. 

Typiflcation: B. spectabilis Hook, ex Hook. f. Type Species (the sole original 
species). 

Caudex subterranean, tuberous, naked, from the crown repeatedly producing one 
to numerous short, slender, more or less determinate frond- and cone-bearing branches 
(sometimes themselves branched); all parts more or less pilose or pubescent when 
young, but glabrescent (except petiole-bases, cataphylls and sporophylls) with maturity. 
Cataphylls short, flat, ovate-triangular, interspersed among the frond-bases. Fronds 
1-several on each slender branch, long-petiolate; decompound by pinnate (or at the 
end sometimes dichotomous) branching of the rhachis, the lowest branches sometimes 
approximate to appear quasi-palmate; rhachis somewhat curved but not twisted, base of 
petiole somewhat thickened, more or less hairy. Pinnules several to fairly numerous 
on each secondary rhachis (which is itself terminally expanded into a pinnule), 
spreading, inserted marginally but towards the adaxial side of the rhachis and more 
or less decurrent, not articulate, simple, entire or lacerate or toothed, obliquely ovate 
or lanceolate, contracted and more or less petiolulate at the base, with close quasi- 
parallel dichotomous venation, stomata confined to the lower surface. Cones of both 
sexes shortly stalked or subsessile, terminal on the short determinate branches, with 
the spirals (parastichies) of sporophylls so arranged that the sporophylls form vertical 
rows (orthostichies) ; sporophylls peltate with expanded oblate-hexagonal terminally- 
compressed spineless ends. Male sporophylls with a very short stipes, a broadly 
obovate-cuneate lamina the proximal half of which has two collateral fertile areas 
on the undersurface, and a more or less hexagonal end compressed to lie in the 
vertical plane. Female sporophylls with a fleshy stipes, the two ovules borne on the 
inward-facing margins ("inner surface") of the expanded hexagonal end. Seeds with a 
fleshy outer coat, inner seed-coat hard. Taproot tuberous, producing apogeotropic as 
well as normal roots. 

Chroinosome number: x = 9 (Darlington and Wylie, 1955). 

Endemic in north-eastern Australia (Queensland), with two species, centred on 
the Cairns and Rockhampton districts respectively, in open spaces in rainforest and in 
eucalypt forest (Text-figure 4). 

Bowenia is a very distinct genus among the Zamiaceae and has not been confused 
with any other. It possesses more advanced features than the other Australian genera 
and appears to be related to such American genera as Zamia, though some of the 
resemblances may be due to parallel evolution or convergence. Bentham's (1873) 
statement that it differs "from Macrozamia only in foliage and in the absence of the 
point to the cone-scales" is not true. The naked caudex, the system of short determinate 
branches, the terminal cone position, and the regular arrangement of the sporophylls 
are all very different from the characters of Macrozaynia, to which Bowenia can be only 
very distantly related. 

The decompound fronds distinguish Bowenia from all other cycad genera. In 
some species of Macrozamia and at times in Stangeria the pinnae are dichotomously 
divided, but the rhachis itself is undivided, so that the fronds may be described as 
simply pinnate with forking pinnae. In Bowenia the rhachis itself branches. 

The fronds are usually described as bipinnate, but Schuster (1932) refers to the 
branching as dichotomous. In Bowenia the frond at maturity may certainly be justly 
called bipinnate; there is a median rhachis (continuous with the petiole) bearing on 
each side two to five lateral secondary rhachides which in turn bear the final segments 
(pinnules) on either side. (The lowest laterals may arise at almost the same point 
due to arrested intercalary growth during development, but the arrangement is still 
pinnate.) Beyond the most distal lateral rhachis the median rhachis itself bears 
pinnules directly (in some cases it forks apparently dichotomously, each branch then 



BY L. A. S. JOHNSON. 



Ill 



bearing pinnules). However, there is little difference between primary and secondary 
rhachides and the junctions tend to be fork-like, though the median branch is the 
larger and, unlike the lateral, continues on to branch again. 

In the light of conditions in other plant groups, this can be regarded as a stage 
in the evolutionary development of a pinnate (or a kind of monopodial) branching 
from dichotomy by the process of "overtopping" (ubergipfelung) . This process must 
be understood, however, in a phylogenetic and not an ontogenetic sense. The frond is 
a determinate branch-system with the whole of its ramifications established in the bud 
stage, and later simply expanded by intercalary growth and cell enlargement. It is 
not strictly comparable with a normal branch-system enlarging by apical growth, 
though probably evolutionarily descended from such a system, as indeed all the fronds 




Text-flg. 4. — Distribution of Bowenia: 1: B. spectahilis ; 2: B. serrulata. 

(leaves) of the various pteropsid lines may be. Similar conditions can be seen in 
various living and extinct fern groups and in the extinct Pteridosperms, and the 
beginnings of the trend are evident even in some of the ancient Psilophytales. 

So far as the Bowenia frond is concerned, the significant point is that the ramifying 
process is less stereotyped than in such groups as Angiosperms or Conifers, so that 
more or less modified dichotomy can occur in various parts of the branching system. 
The same applies to other cycads, but usually less strikingly. In Bowenia the capacity 
for dichotomy and for overtopping has made possible a particular kind of frond 
specialization apparently suited to the environment. Bowenia is not a primitive cycad, 
and we need not assume that the decompound frond has been continuously retained 
during its evolutionary history. It may well be a secondary acquisition, which would 
remain possible so long as the mechanism for unstereotyped branching was not lost. 

Both species have been cultivated to some extent as ornamentals, but are rather 
tender even in the Sydney climate. 



Key to the Species. 

1. Pinnules entire or a few of them irregularly lacerate. Caudex elongate, with 1-5 frond- 
bearing branches. N.E. Queensland 1. B. spectabilis. 

3.* Pinnules regularly serrate. Caudex subspherical, with 5-20 frond-bearing branches. Central 
eastern Queensland 2. B. semilata. 



112 THE FAMILIES OF CYCADS AND THE ZAMIACEAE OF AUSTRALIA, 

1. B. spECTABiLis Hook, ex Hook. £. ia Bot. Mag., 89 (1863), sub t.5398; ihid., 98 (1872), 
sub t.6008; F. Muell., Fracjm. PMjtogr. Austral., 5 (1866), 171; A.DC, Prodr., 16, 
pt. 2 (1868), 524; Miq. in Arch. Neerl., 3 (1868), 254, and in Versl. Meded. Koninkl. 
ATcad., ser. ii, 3 (1869), 57; Benth., F^l. Austral., 6 (1873), 254; Regel in Acta Hort. 
Petrol).. 4 (1876), 316; F. M. Bail., Queensl. Fl., 5 (1902), 1507, in part, excl. var.; 
Chamberlain in Bot. Gaz., 54 (1912), 419; Domin in Bibl. Bot., 20 (1915), 244; 
Schuster, Pflanzenr., IV, i (1932), 85, in part, excl. var. 

Typification: J. D. Hooker mentioned a specimen of A. Cunningham's from the 
Endeavour River, 1819, but stated that the description was made from a living plant 
with dried leaves and a male cone, from Rockingham Bay, sent by W. Hill to Kew in 
1863. This material is thus the Holotype and is represented by Bot. Mag., t.5398. 

Caudex subterranean, 2-10 (or more?) cm. diam., elongate, passing into the elongate 
tuberous tap-root, its crown bearing (at any one time) 1-3 (-5) short slender frond- 
and cone-bearing branches. Fronds few (ca. 1-7), erect, with 4-10 more or less spreading 
branches (pinnae), to 100-200 cm. long (the flrstiones of a new series often smallei, 
from 40 cm.) and 100 cm. broad, somewhat pilose when young, but glabrescent with 
age; petiole ca. half the total length of the frond, slender (ca. 2-8 mm. thick), almost 
terete, but slightly channelled above, somewhat thickened and more or less villous at 
the base, sometimes also with shorter stiff hairs persistent throughout the proximal 
half; primary and secondary rhachides slender, slightly flattened but 1-2-channelled 
above and laterally ridged from the decurrent pinnule-bases, rounded to subangular 
beneath. Pinnules 7-30 on each pinna (also on the terminal unbranched portion of 
the primary rhachis), spreading, entire or a few with occasional coarse lacerations, 
rather thin and flexible, shining above, obliquely falcate-lanceolate (or more or less 
ovate on small fronds), 7-15 cm. long, 1-5-4 cm. broad, with numerous more or less 
parallel dichotomously-branched nerves visible on both surfaces but not prominent, 
tapered to the acute or often acuminate non-pungent apex, gradually contracted and 
finally more or less petiolulate at the slender decurrent base. Cones not seen mature, 
the following is partly derived from descriptions by other authors: Male cones shortly 
stalked, ovoid, to 5 cm. long and 2-5 cm. diam., sporophylls broadly obovate-cuneate 
with dilated, truncate, subhexagonal, more or less tomentose ends. Female cones sub- 
sessile, oblong-globose and rounded at the apex, to ca. 10 cm. long (or longer?), 7-10 cm. 
diam.; sporophylls about 8-stichous, the expanded ends oblate-hexagonal (in the 
vertical plane), 3-5-5 cm. broad, ca. 1-5 cm. high, terminally more or less truncate- 
pyramidal with a central depression (said to be more or less tomentose but glabrescent 
in the immature example seen). Seeds to 3-2 cm. long, 1-8 cm. thick. 

Chromosome number: 2n = 18. 

Bistrtt)Ution: Queensland: North-eastern coast and ranges from Cooktown to 
Rockingham Bay district, in the more open situations in and around rainforest. 

Specimens examined: Queensland: Whyanbeel Creek, 7 miles north of Mossman, 
M. Tindale, 15.vii.l957 (NSW.42280); Cape Tribulation, W. W. Mason, i.l947 
(NSW.40615); Yarrabah Mission, Trinity Bay near Cairns, P. R. Messmer, 16.vii.l952 
(NSW.30535); Jordan's Creek, P. R. Messmer, ll.viii.l954 (NSW.30537); Babinda, 
W. W. Watts, vii.1913 (NSW.40618); Atherton district, per H. L. White, 1912 
(NSW.40620); Malanda, C. T. White, i.l918 (NSW.40617); Geraldton, Johnstone River, 
S. W. Jackson, 1908 (NSW.40619); Rockingham Bay (NSW.40616). 

The concept of B. spectahilis has been extended by some authors, including 
Schuster (1932), to include B. serrulata, but the latter differs in a number of characters 
and the two population-groups seem quite isolated (see below, under B. serrulata). 
According to Chamberlain (1912), B. spectabilis does not form dense stands like those 
of B. serrulata, and each plant produces fewer fronds. The species has been cultivated 
to some extent. 



BY L. A. S. JOHNSON. 113 

2. B. SERRULATA (W. Bull) Chamberlain in Bot. Gaz., 54 (1912), 419. 

Basionym: B. spectaMUs Hook, ex Hook. f. [var.] serrulata W. Bull, Catal. (1878). 
4, t.5.* [Chamberlain gave the citation "(Andre) Chamberlain, n. comb." and referred 
to B. spectaMUs Hook. f. var. serrulata Andre, III. Hort.. 26 (1879), 184, t.366. However, 
Andre did not publish this as a new variety of his own; he cited it as "Hort. Angl." and 
his plate is identical with Bull's original, though he made no explicit reference to Bull. 
The same plate, with similar descriptions, and references to Bull's Catalogue, was 
reproduced also by T. Moore in Florist and Pomologist (July, 1878), 107, as "Boioenia 
spectaMUs serrulata". as well as in Gartenflora, 27 (1878), 314, and in III. Gartenz. 
Stuttgart, 23 (1879), 99, t.l5, as B. spectaMUs var. serrulata. Since Bull's was the 
original valid publication of the variety from which the others were derived, it should 
be cited as the basionym. Despite the custom of citing the author of the variety in 
parentheses, the specific name B. serrulata dates, for purposes of priority, only from 
1912.] 

Typification: The description was taken from a living (vegetative) plant in Bull's 
collection in London; his plate must serve as Hot.otype. It unmistakably represents 
the present species. 

Synonymy: B. spectaMUs Hook, ex Hook. f. var. serrata F. M. Bail., Syn. Queensl. 
Fl. (1883), 501; Queensl. Fl.. 5 (1902), 1507. 

In most respects similar to B. spectaMUs, but differs as follows: Caudex subter- 
ranean, subspherical, to 20-25 cm. (or more) diam., with 5-20 short slender frond- 
and cone-bearing branches. Fronds (of well-developed plants) ca. 5-30, the pinnule^ 
sharply serrate (except in the lower J-J) with rather pungent teeth 1-3 mm. long, 
sometimes also a few of them coarsely lacerate as (occasionally) in B. spectaMUs. 
[There may be some difference in the cones, of which I have no material, but none is 
recorded.] 

Chromosome number: 2n - 18. 

Distribution: Central eastern Queensland, in the vicinity of Rockhampton, iu 
oucalypt (dry sclerophyll) forest. 

Specimens examined: Queensland: Byfield, Busch, x.1916 (NSW. 40621); Byfleld, 
O. D. Evans, iv.l924 (SYD). Described also from cultivated material. 

Though it has been treated by many authors, including the most recent mono- 
grapher (Schuster, 1932), as a variety of B. spectaMUs, B. serrulata differs consistently 
from the northern plants, as pointed out by Chamberlain (1912), in the serration of 
the pinnules (a condition quite distinct from the coarse laceration found in both 
species), the subspherical shape of the caudex and the more numerous frond-bearing 
branches. These characteristics are maintained in cultivation, given suitable conditions 
for full development (in pots, B. serrulata often grows poorly and produces few 
fronds). Furthermore, the populations are quite isolated in nature and occupy rather 
different ecological niches (unlike B. spectabilis, B. serrulata grows in rather dry 
sclerophyll forest and forms fairly dense and extensive stands). They thus agree 
v/ith any reasonable concept of specific distinction, since each is now evolutionarily 
independent and differs materially from the other. 

B. serrulata is sometimes grown in gardens and conservatories for its ornamental 
fronds. 

Index of Names. 

Names accepted in this revision (Part II only) appear in small capitals, with reference 
to their numbers in the formal section (I, 1, etc.). Valid synonyms, in italic, are referred 
to their correct nomenclatural position (in roman) as determined by the identity of their types. 
Invalid nomina nuda and misapplications are not cited (for Schuster's misapplications see 
Table 1, p. 74). All epithets are decapitalized and terminations in "-i/-ii" are given in the 

* I am indebted to Dr. R. Melville of the Royal Botanic Gardens, Kew, for copies of the 
descriptions in Bull's Catalogue and the Stuttgart Gartenzeitung. In Index Londinensis the 
entry in the latter publication is wrongly given as "B. serrata". In fact, it appeared as 
"Bowenia spectabilis. Hooker, var. serrulata". Previous to Bull's valid publication the 
trinomial appeared in Gard. Ghron., n.s., 8 (1877), 310, as a notnen nudum. 



114 THE FAMILIES OF CYCADS ANI) THE ZAMIACEAE OF AUSTRALIA, 

correct form as authorized by the I. C.B.N. (1956), irrespective of the original. Other mis- 
spellings are given in quotation marks. 
BovirBNiA Hook, ex Hook. f. (1863). III. 

BowENiA SERRULATA (W. Bull) Chamberlain (1912). Ill, 2. 
B. SPECTABiLis Hook, ex Hook. f. (1863). III. 1. 
h. spectabilis var. serrata P. M. Bail. (1883) = B. serrulata. 
B. spectaMUs [var.] serriclata W. Bull (1878) = B. serrulata. 
Catakidozamia W. Hill (1865) = Lepidozamia. 
Catakidozamia hopei W. Hill (1865) = Lepidozamia hopei. 
Cycas riedlei Pisch. ex Gaudich. (1826) = Macrozamia, riedlei. 
Encephalartos sect. Lepidozamia (Kegel) Miq. (1863) = Lepidozamia. 
Eiicephalartos sect. Macrozamia Miq. (1863) = Macrozamia sect. Macrozamia. 
Encephalartos sect. Parazamia Miq. (1863) = Macrozamia sect. I'arazamia. 

Encephalartos denisonii (C. Moore and P. Muell.) P. Muell (1859) - Lepidozamia peruffskyana. 
E. donglasii P. Muell. (1883) = Macrozamia miquelii. 
E. dyeri P. Muell. (1885) = Macrozamia riedlei. 
E. fraseri (Miq.) Miq. (1863) = Macrozamia riedlei. 

E. macdonnellii P. Muell. ex Miq. (1863), (sphalm. ■• macdoiieUi" ) = Macrozamia macdonnellii. 
E. miquelii P. Muell. (1862) = Macrozamia miquelii. 

E. moorei (P. Muell.) P. Muell. (August, 1881) = Macrozamia nioorei. 
E. oldfteMii Miq. (1863) = Macrozamia riedlei. 
E. pauli-cjxiilielmi (W. Hill and P. Muell.) P. Muell. (1859) = Macrozamia pauli-guilielmi ssp. 

pauli-guilielini. 
E. preissii (Lehm.) P. Muell. (1859) = Macrozamia riedlei. 
E. spiralis (Salisb. ) Lehm. (1834) = Macrozamia spiralis. 
E. spiralis var. diplomera P. Muell. (1866) = Macrozamia diplomera. 
E. spiralis var. major Miq. (1863) = Macrozamia iniquelii. 
Lepidozamia Kegel (1857). I. 

Lepidozamia denisonii (C. Moore and P. Muell.) Kegel (1875) = L. peroffskyana. 
Lepidozamia hopei Kegel (1876). I, ]. 

[I>. HOPEITES (Cookson) L. .Johnson (1959), ,s-;j. foss. See under 1, 1.] 
L. PEROFFSKYANA Kegel (1857). 1. 2. 
Macrozamia Miq. (18 42). II. 

Macrozamia sect. Genuinae Miq. (1868), Jio»i. illegit. = M. sect. Macrozamia. 
Macrozamia sect. Lepidozamia (Kegel) Miq. (1868) = Lepidozamia. 
Macrozamia sect. M.\crozam\a. II. i. 

Macrozamia sect. Monooccidentales Schuster (1932), nom. illefjit. = M. sect. Macrozamia. 
Macrozamia sect. Monoorientales Schuster (1932) iiotn. illegit. = Lepidozamia. 
Ma(3Kozamia sect. Parazamia (Miq.) Miq. (1868). II, ii. 

Macrozamia sect. Polyorientales Schuster (1932), nam. illegit. = M. sect. Macrozamia. 
Macrozamia sect. Polyorientales subsect. Acu.tae Schuster (1932) = M. sect. Parazamia. 
Macrozamia sect. Polyorientales subsect. Attenuatae Schuster (1932), nom. illegit. = M. sect. 

Macrozamia. 
Macrozamia sect. Polyorientales subsect. Curvatae Schuster (1932) = M. sect. Parazamia. 
Macrozamia communis L. Johnson (1959). II, 5. 
M. corallipes Hook. f. (1872) = M. spiralis. 
M. cylindrica C. Moore (1884) = M. miquelii. 

M. denisonii C. Moore and P. Muell. (18 58) = Lepidozamia peroffskyana. 
M. deniso7iii var. hopei (W. Hill) Schuster (1932) = Lepidozamia hopei. 
M. diplombea (P. Muell.) L. Johnson (1959). II, 6. 
M. douglasii W. Hill ex P. M. Bail. (1883) = M. miquelii. 
M. dyeri (P. Muell.) C. A. Gardn. (1930) = M. riedlei. 
M. i-AwcETTii C. Moore (1884). II, 14. 

M. flexuosa C. Moore (1884) = M. pauli-guilielmi ssp. flexuo.sa. 
M. fraseri Miq. (1842) = M. riedlei. 
M. heteromera C. Moore (1884). II, 11. 

M. heteromera var. dicranophylloides Schuster (1932) = M. stenomera. 
M. heteromera var. glauca C. Moore (1884) = M. heteromera. 
M. heteromera var. tenuifolia C. Moore (1884) = M. stenomera. 
M. heteromera var. tenuifolia Schuster (1932), nom. illegit. = M. stenomera. 
M. heteromera var. tenuifolia i. harmsii Schuster (1932) = M. heteromera. 
M. hopei W. Hill ex P. M. Bail. (1886) = Lepidozamia hopei. 
[ikf. hopeites Cookson (1953), sp. foss. = Lepidozamia hopeites.] 
M. LUCID A L. Johnson (1959). II, 8. 

M. MACDONNELLII (P. Muell. ex Miq.) A. DC. (1868). II, 3. 
M. mackenzii Hort. ex Mast. (1877) = M. miquelii. 
M. macleayi Miq. (1868) = M. miquelii. 
M. MIQUELII (P. Maeil.) A.DC. (1868). II, 4. 
M. MOOREI P. Muell. (March, 1881). II, 1. 



BY L. A. S. JOHNSON. 115 

M. moiiiUperriensis F, M. Bail. (1886) = M. miquelii. 

M. oUlfleldii (Miq. ) A.DC. -(1868) = M. riedlei. 

M. PAULl-GUiLiELMi "W. Hill and F. Muell. (1859). II, 13. 

M. PAULi-GUiLiELMi ssp. FLExuosA (C. Moore) L. John.son (UTiy). II, 13c. 

M. PAULI-GUILIBLMI SSp. PAULI-GUILIBLMI. II, 13a. 

M. PAULi-GuiLiELMi ssp. PLURiNEEViA L Johnson (19 59). II, 13b. 

M. peroffskyana (Kegel) Miq. (1868) = Lepiclozamia peroffskyana. 

M. PLATYHACHIS F. M. Bail. (1898). II, 7. 

M. i)lnmosa Hort. ex Mast. (May, 1875) = (probably) M. pauli-guilielmi ssp. pauli-guilielmi. 

M. phimosa Hort. ex auct. anon. (June, 1875) = M. pauli-guilielmi ssp. pauli-guilielmi. 

yi. yi-eissU Lehm. (1S44) = M. riedlei. 

M. preissii ssp. dyeri (F. Muell.) Schuster (1932) = M. riedlei. 

M. RIEDLEI (Fisch. ex Gaudicli.) C. A. <Jardn. (1930), (-reidleV) . II, 2. 

M. SBCUNDA C. Moore (1884). II, 10. 

.1/. secunda var. dichotoma C. Moore and Betche (18 93) = (probably) M. heteroniera x M. 

secunda. 
M. SPIRALIS (Salisb. ) Miq. (1842). II, 9. 

M. spiralis var. IcoraUipes (Hook, f . ) Benth. (1873) => M. spiralis. 
M. spiralis var. Tcylindracea Benth. (1873) = M. pauli-guilielmi (ssp.?). 
M. spiralis var. oylindiica Kegel (1876), nom. dub. = M. miquelii or M. eomnuinis. 
M. spiralis var. cylindrica (C. Moore) Maid, and Betche (1916), nom. illetjit. - M. miquelii. 
M. spiralis! var. diqilomera (F. Muell.) A.DC. (1868) = M. diplomera. 
M. spiralis var. ''.diplomera F. Muell. ex Benth. (1873) = M. diplomera. 
M. spiralis var. eximia Kegel (1876) nom. duh. = M. miquelii or M. communis. 
71/. spiralis var. fawcettii (C. Moore) Maid, and Betche (1916) = M. fawcettii. 
M. spiralis var. flexuosa (C. Moore) Maid, and Betche (1916) = M. pauli-guilielmi ssp. tlexuosa. 
M. spiralis var. frasieri Kegel (1876), notn. duh. = M.' miquelii or M. communis. 
M. spiralis var. lieteromera- (C. Moore) Maid, and Betche (1916) = M. heteromera. 
ill', spiralis var. hillii Kegel, (1876), noin. duh. = M. miquelii or M. communis. 
-M. spiralis var. Isecunda Benth. (1873) = M. secunda. 

M. spiralis var, seomda (C. Moore) Maid, and Betche (1916) = M. secunda. 
M. STENOMERA L. .lohnson (1959). II, 12. 

IM. tridentata (Willd.) Kegel. See list of excluded and dubious names, below.] 
ilf. trklentata ssp. cylindricai (C. Moore) Schuster (1932) = M. miquelii. 
ilf. tridentata ssp. cylindrica var. corallipes (Hook, f . ) Schuster (1932) =i M. spiralis. 
M. tridentata ssp. cylindrica- var. corallipes f. dielsii Schuster (1932) = M. spiralis. 
M. tridentata ssp. cylindrica var. corallipes f. vavilovii Schuster (1932), nom. duh. = M. 

miquelii or M. communis. 
M. tridentata ssp. cylindrica var. corallipes f. wallsendensis Schuster (1932) = M. communis. 
[M. tridentata ssp. cylindrica var. punyeiis (Ait.) Schuster (1932), nom. duh. (quoad 

typ.) = Encephalartos sp. ?] 
M. tridentata ssp. cylindrica. var. punyens f. diplomera (F. Muell.) Schuster (1932) = M. 

diplomera. 
M. tridentata ssp. cylindrica var. piDifjens f. hillii Schuster (1932), nom. duh. = M. miquelii or 

M. communis. 
M. tridentata ssp. cylindrica var. secunda (C. Moore) Schuster (1932) = M. .secunda. 
M. tridentata ssp. mountperriensis (F. M. Bail.) Schuster (1932), ("mountpei~ryensis"), = M. 

miquelii. 
M. tridentata ssp. inountpenriensis var. donylasii (F. Muell.) Schuster (1932) = M. miquelii. 
M. tridentata ssp. mountperriensis var. macken.zii (Hort. ex Mast.) Schuster (1932) = M, 

iniquelii. 
M. tridentata ssp. mountperriensis var. miquelii (F. Muell.) Schuster (1932), nom. illefjit. = M. 

miquelii. 
III. tridentata ssp. mountperriensis var. miquelii f. milkaui Schuster (1932), liojit. illegit. = M. 

miquelii. 
M. tridentata ssp. mount perriensis var. miquelii f. ohlonijifolia (Kegel) Schuster (1932) = M. 

miquelii. 
M. tridentata var. oblonyifolia Kegel (1876) = M. miquelii. 
Xamia spiralis Salisb. (1796) = Macrozamia spiralis. 

Excluded and Dubious Names. 

This list includes only names which are wholly dubious, possibly not being based on 
Australian plants at all. Nan)es which can be placed to within a couple of species are 
included in the Index, above. 

Encephalartos pungens (Ait.) Lehm., nom. duh. = Encephalartos sp.? See under II, 5 and 9. 
E. tridentata (Willd.) Lehm., nom,. dub. = Macrozamia or Encephalartos sp. ? (juvenile). See 

under II, 4, 5 and 9. 
Macrozamia tridentata (Willd.) Kegel, nom. dub. (quoad basionym.) = Maci'ozamia or 

Encephalartos sp. (juvenile). See under II, 4, 5 and 9. 



116 ' THF: families of C'YCADS and the ZAMIACEAE of AUSTRALIA, 

M. tridentatd ssp. cyllndrica var. -pungenR (Alt.) Schuster, nmn. dnh. (quoad 

l:)asionym.) = Enccphalartos sp.? See under II, 5 and 9, 
Zwiiiia pallida Salisb., nom. dul). = Enoephalartos or Macrozamia, sp.? (juvenile). See under 

II, 5. 
7,. pungens Alt., nom. dnh. = Enoephalartos sp.? See under II, 5 and 9. 
'/,. tridentatd Wllld., nom. dub. = Jllacrozaviia or Enoephalartos sp. ? (juvenile). See under II, 

4. 5 and 9. 

Acknoioledgements. 

My thanks are due to Mr. R. H. Anderson, Chief Botanist and Curator, National 
Herbarium, Royal Botanic Gardens, Sydney, for making this w^ork possible, and to the 
past Directors of the Melbourne and Brisbane Herbaria for the opportunity to examine 
several specimens. Dr. Joyce Vickery has given generously of her time and experience 
in the discussion of some of the difficult matters of typification and nomenclature 
which abound in Macrozamia; responsibility for the decisions, however, rests with 
myself. Dr. R. Melville of Kew has been most helpful in sending me copies of obscure 
references. Miss D. P. Ragg, Librarian at the National Herbarium, has shown great 
patience and enterprise in obtaining many rare publications. 

Dr. Isabel Cookson freely discussed with me her work on frond anatomy in 
Macrozamia and Leindozamia. and some of her unpublished findings, though not 
specifically mentioned in this revision, have helped to influence and confirm my 
judgements. But for Mr. G. Chippendale, Botanist, Animal Industry Division, Alicf 
Springs, I should not have been able to examine Macrozamia macdonvellii in the field. 
Mr. N, Forde had earlier sent me excellent material and photographs of this species. 
Finally, I thank all of my colleagues at the National Herbarium for co-operation in 
various ways, and particularly Mr. B. F. Constable, Official Collector, for his efforts and 
companionship in the field, especially in the re-discovery of the magnificent M. moorel 
in New South Wales. 

References. 
Anderson, R. H. (Ed.), in press — Flora of Neiv South Wales. Pts. 1-lS. Contrih. N.S.W. 

Nat. Herbariunij Flora series. Government Printer, Sydney. 
Arnold, C. A., 1948. — Classification of gymnosperms from the viewpoint of paleobotany. Bot. 

Gaz., 110 : 2-12. 
Bailbt, F. M., 1902. — The Queensland Flora, 5. Queensland. Government , Brisbane. 
Bbntham, G., 1873. — Flora AustraUeiisis, 6. Reeve, London. 
Brough, p., and Taylor, M. H., 1940. — An investigation of the life cycle of Macrozamia 

spiralis Miq. Proc. Linn. Soc. N.S.W., 65: 494-524. 
Brown, R., 1810. — Prodronius Florae Novae Holla/ndiae et Insulae Van-Dietnen. Johnson, 

London. 
Candollb, a. De, 1868. — Cycadaceae.. In Prodromns systematis natnraJis regni vegetabilis, 

16, Pt. 2 : 522-548. Masson. Paris. 
Chamberlain, C. J., 1909. — Dioon spinnlosum. Bot. Gaz., 48 : 401-413. 

, 1912. — Two species of Bowenia. Bot. Gaz., 54 : 419-423. 

, 1913. — Macrozamia moorei, a connecting link between living and fossil cycads. Bot. 

55: 141-154. 

1919. — The Living Cycads. TJniverrsity of Chicago Press, Chicago. 

I. C, 1953. — On Macrozamia hopeites — an early Tertiary cycad from Australia. 

Phytomorphology, 3 : 306-312. 
Cronquist, A., 1957. — Outline of a new system of families and orders of Dicotyledons. Bxdl. 

Jard. Bot. Brux., 27: 13-40. 
IJAULINGTON, C. D., and Wylie, A. P., 1955. — Chromosome Atlas of Flowering Plants. Allen 

and Univin, London. 
Davis, G. L., 1952. — Revision of the genus Calotis R.Br. Proc. Linn. Soc. N.S.W., 77 : 146-188. 
Eambs, a. J., 1952. — Relationships of the Ephedrales. Phytomorphology , 2 : 79-100. 
, 19 53. — Floral anatomy as an aid in generic limitation. In Plant genera, their nature 

aiiA definition (Symposium). Chron. bot., 16 : 126-132. 
Florin, R., 1948. — On the morphology and relationships of the Taxaceae. Bot. Gaz., 110: 31-39. 
, 1954. — The female reproductive organs of conifers and taxads. Biol. Rev., 29: 367- 

389. 
Fkitsch, F. E., 1945. — The Structure and Reproduction of the Algae, 2. Cambridge University 

Press, Cambridge. 
Gardner, C. A., and Bennetts, H. W., 1956. — The Toxic Plants of Western Australia. W.A. 

Newspapers, Perth. 
Hooker, J. D., 1863. — Boivenia spectabilis. Curtis's Bot. Mag.. 89 : under tab. 5398. 



BY L. A. S. JOHNSON. 117 

HoOKEBj J. D., 1872. — Bouenia spectabilis. Curtis's Bot. Mag., 98 : under tab. Ij008. 

Hurst, E., 1942. — The Poison Plants of Nexo Sotith Wales. N.S.W. Poison Plants Committee^ 

Sydney. • 
Hutchinson, J., 1924. — Contributions towards a phylogenetic classification of flowering' plant., : 

III. The genera of Gymnosperms. Ketv BxM.. 1924 : 49-66. 
Johnson, L. A. S., in press. — Part I. Zamiaceae. In Flora of Neiv So-iith Wales, Pts. 1-18. 

Ed. R. H. Anderson. Contrib. N.S.W. Nat. Herharium, Flora series. Government Printer, 

Sydney. 
Lanjouw, J. (Ed.), 1956. — International Code of Botanical Nomenclature. International 

Bureau for Plant Taxonomy and Nomenclature, Utrecht. 
Light, K. E., 1924. — The ovule and development of the female gametophyte of Macro^at>iia 

Fraseri. Aim. Bot., 38: 337-349. 
Maiden, J. H., and Betche, E., 1916. — Coisns of Neiv South Wales Plants. Government 

Printer, Sydney. 
Melville, R., 1957. — Encephalartos in Central Africa. Keic Bull., 1957 : 237-257. 
MiQUEL, F. A. G. [F. A. W.], 1842. — Monographia Cycadiearum. Natan, Trajecti ad Rhenuin. 

, 1861. — Prodromns systeinatis cycadearuni. Van der Post, Utrecht. 

, 1863. — Over de Cycadeen in Nieuw-Hclland. Versl. Meded. koninkl. Akud. 

Wetenschappen, 15 : 363-376. 
, 1868. — Nouveaux materiaux pour servir a la connaissance des Cycadees. Arch. 

neerl. Sci. exact, nat., 3 : 193-254. 
, 1869. — Nieuwe Bijdragen tot de kennis der Cycadeen. Versl. Meded. koninkl. Akad. 



Wetenschappen, (2nd s. ) 3: 1-57. 
MooRB, C, 1884. — Notes on the genus Macrozamia. J. Roy. Soc. N.S.W.. 17 : 115-122. ; 
— — — , and Behtche, E., 1893. — Handbook of the Flora of New South Wales. Government 

Printer, Sydney. 
PlLGER, R., 1926. — Cycadaceae. In Die Natiirlichen Pflanzenfamilien, 2 Aufl., 13 : 44-82. Ed". 

A. Engler. Engelmann, Leipzig. 
Plumstead, E. D., 1956. — Bisexual fructifications borne on Glossopteris leaves from South 

Africa. Palaeontographica , 100 : 1-25. 
Rbgel, E., 1857. — Zwei neue Cycadeen. B^al. Soc. imp. nat. Mosc, 1857, 163-191. 

, 1876. — Cycadearum generum specierumque revlsio. Acta Hwt. Petrop., 4 : 275-320. 

Reiohbnbach, H. G. L., 1837. — Handbuch des natiirlichen Pflanzensystenis. Arnold, Dresden. 
Schuster, J., 1932. — Cycadaceae. In Das Pflanzenreich, IV, i: 1-168. Ed. A. Engler. Engelmann, 

Leipzig. 
Simpson, G. G., 1945. — The principles of classification and a classification of mammals. Butt. 

Amer. Mus. Nat. Hist., 85: 1-xiv, 1-350. 

■ — , 1953. — The Major Features of Evolution. Cohimbia University Press, New York. 

Specht, R. L., 1958. — An introduction to the ethno-botany of Arnhem Land. Rec. Avier.-Aus-t. 

Exped. Arnhem Land, 3:479-503. 
Stbarn, W. T., 1941. — Bibliographical notes. CXII. Dates of publication of De Candolle's 

'Systema' and 'Prodromus'. J. Bot., 79: 25-27. 
Stebbins, G. L., 1950. — Variation and Evolution in Plants. Oxford University Press, London. 
Thieeet, J. W., 1958. — Economic botany of the cycads. Econ. Bot., 12: 3-41, 
Thomas, H. H., and Bancroft, N., 1913. — On the cuticles of some recent and fossil cycadean 

fronds. Trans. Linn. Soc. London, Ser. 2, Botany, 8 : 155-204. 
Webb, L. J., 1948. — Guide to the medicinal and poisonous plants of Queensland. Co-unc. Sci. 

Ind. Res. Bull., 232 : 1-202. 
WEnrrSTBiN, R. von.. 1923. — Handbuch der systematischen Botanik, 3 Aufl., 1. Deuticke, 

Leipzig'. 



118 



ROOT PARASITISM IN ATKlNSONiA. LIGUHTRINA (A. CUNN. EX F. MUELL.) 

F. MUELL. 

By Bakkaha P. Mioa'ziks, Department of Botany, University of Sydney, and H. S. McKee, 

Joint Plant Physiology Unit, Department of Botany, University of Sydney, and Division 

of Food Preservation and Transport, C.S.I.R.O. 

(Fourteen Text-figures.) 

[Read 29th April, l!ir,9.] 



Synoiisis. 
A brief survey i.s given of the terrestrial Loranthaceae. The structure of haustoria of 
Atkinsonia liyustrhia (A. Cunn. ex F, Muell. ) F. Muell. is described and compared with that 
found in some other phanerogamous root parasites. 



Inteoduction. 

The family Loranthaceae contains several hundred species, nearly all being 
parasitic shrubs growing on the branches of trees. Terrestrial species include Nuytsia 
floribunda R.Br., the Christmas tree of Western Australia, several South American 
species of Gaiadendron G. Don, Lorcmthus terrestris Hook. f. and L. Iigustri7ius Wall, 
in India, L. huchneri Engl, in Africa and Atkinsonia ligustrina, which occurs sporadic- 
ally on dry rocky ridges in a restricted area of the Blue Mountains of New South 
Wales. Various authors, including Danser (1933), Moore and Betche (1893), Ruiz 
(1940) and Skene (1924), state that some or all of these terrestrial species are non- 
parasitic, though it should be remembered that with the older writers "non-parasitic" 
may merely mean "non-epiphytic". Hooker (1890) stated that Loranthus ligustrinun 
is "apparently terrestrial, a root parasite?", and referred to L. terrestris as follows: 
"always growing from the ground and hence probably a root parasite. Dr. Thomson 
and I gathered this plant at seven or eight localities and never found it attached to an 
aerial tree branch." Welwitsch (Hiern, 1900) found L. hiichneri "growing parasitically 
on the roots of a Malvacea ( probably a species of ^ida ) , though at first sight terrestrial 
and not parasitical". 

Herbert (1919) has shown that Nuytsia florihunda is a root parasite, as was indeed 
con.iectured by Harvey in 1854 (Anonymous, 1869). It attacks a wide range of hosts, 
including both native and introduced species. The question whether the species of 
Gaiadendron are parasitic or not seems to have been little studied. Ruiz (1940) 
describes Loranthus luteus, presumably one of the species now referred to Gaiadendron, 
as a tree ten to twelve yards high, very beautiful on account of its abundance of yellow 
flowers, and not parasitic. MacBride (1937) cites in this genus two species referred to 
Loranthus by Ruiz and Pavon in 1802, and to Phrygilanthus in 1868 by Eichler. These 
species, Gaiadendron elUpticum (Ruiz et Pavon) Baehni ex MacBride and G. lanceolatuni 
(Ruiz et Pavon) Baehni ex MacBride, he states to be "as far as is known, parasitic 
shrubs". Gaiadendron, on this interpretation, appears to contain both terrestrial and 
epiphytic species. One of the former, G. punctatum (Ruiz et Pavon) G. Don, described 
as a tree over 20 m. high, seems to be the largest of the known Loranthaceae. Finally, 
MacBride (1937) refers, rather cryptically, to Phrygilanthus eugenioides (H.B. et K.) 
Eichler (= G. eugenioides G. Don) as "half-parasitic on trees or growing independently". 
A personal communication from Dr. M. Cardenas of Cochabamba, Bolivia, mentions 
this species as growing either in the soil or as a branch parasite. 

Amyema scandens (Van Tieghem) Danser seems able to grow independently in 
the ground for a time before attacking a host plant. A specimen (McKee 2375) of this 

Proceedings of the Linnean Society of New South Wales. 1959, Vol. Ixxxiv, Part 1. 



1!Y JiAKBARA P. MENZIES AND H. S. MCKEE. 119 

Species collected In New Caledonia had a stem two cm. in diameter, attached by many 
haustoria to the trunlt of the host (a small tree of Spermolepis gummifera Brong. et 
Gris). Near ground level the parasitic stem became much thinner and separated 
from the host, entering the soil where it bdre what appeared to be a poorly developed 
root system. The underground parts and the narrow stem immediately above them 
were dead, though still attached to the living parasitic stem higher up the stem of 
the host. Amyenia scandens grows to a large plant climbing to the tops of tall trees, 
and in such cases may show no indication that it began life in the ground. Similar 
observations on this species, and on Amylotheca pyramidata (Van Tieghem) Danser, 
liave been recorded by other collectors in New Caledonia, e.g. Compton (Rendle, Baker 
and Moore, 1921: see description of Loranthus c/laucescens S. Moore, a name treated by 
Danser (1936) as a synonym of Amyema scandens) and Virot (Guillaumin and Virot, 
1953; Virot, 1956). It is of interest that these species seem capable of limited indepen- 
dent growth in the soil; they have not been adequately studied, but there is no evidence 
that they are ever root parasites. Later collectors have confirmed but hardly extended 
the following notes made by Balansa on the type specimen (Balansa 498) of Neophyluni 
scandens Van Tieghem (now Amyema scandens (Van Tieghem) Danser), which he 
collected in New Caledonia in 1869: "pourvue d'une tige ayant ses racines dans le sol 
meme, elle grimpe au sommet des plus grands arbres, en implantant dans leur bois de 
distance en distance une racine qui egale souvent en grosseur la tige mere." 

Taxonomy and Nomenclature. 

The species which is the subject of this paper has a somewhat involved synonymy. 
It was discovered by Allan Cunningham in 1817. He named it Nuytsia ligustrina, the 
name being first published by Lindley (1839), who, however, gave no description. 
The first description was that of Mueller (1861) who retained Cunningham's name, but 
later (Mueller, 1865), having received fruiting material, removed the species from 
Nuytsia and erected a new genus Atkinsonia for it, named after Miss L. Atkinson, who 
collected extensively in the Blue Mountains. To quote Mueller's own words: ". . . dicavi 
praeclarae Ludovicae Atkinson, cujus henignitate hanc et multas alias Neo-Gambriae 
plantas deheo." The genus has had a chequered history. Bentham and Hooker (1883) 
sank it in Loranthus. while Engler (1894) referred the single species to Gaiadendron. 
Later (Engler, 1897), he divided that genus into two sections: Eugaiadendron, con- 
taining the South American species, and Atkinsonia with one Australian species. 
Finally, Engler and Krause (1935) restored Atkinsonia as a separate genus, the dis- 
tinguishing character being that in Gaiadendron the flowers are in racemes of triads 
and in Atkinsonia in simple triads. Two names cited by Jackson (1895), apparently 
as distinct, may refer to the present species. These are Loranthus atkinsonae Berith. 
and L. epigaeiis F. Muell. Mueller (1865) cites the latter as a synonym, but seems not 
to have published it; Danser (1933) cites it as L. epigaeus Jacks. The specific name 
Hgustrinus is preoccupied in Loranthus by L. ligustrinus Wall., an Indian species. 
The present species is figured (as L. atkinsoniae Benth. ) by Oliver (1880) and by 
Blakely (1922, Plate XXVIII). The only previously published investigation of its 
parasitism, except for a preliminary note (McKee, 1952) by one of the present authors, 
appears to be due to Blakely (1922), who examined the roots of several plants without 
finding haustoria. 

DiSTKlBUTlON AND HoST RANGE. 

The species is known only from the Blue Mountains of New South Wales, where it 
occurs sporadically, often on dry rocky ridges. It is not common, but is usually found 
in groups of from two to twenty plants. A suggestion that it occurs outside the Blue 
Mountains comes from Mueller (1860), who states that it "occurs in the Blue Mountains 
and is said to be found also on the north-eastern tributaries of the Darling". This 
suggests a northwards extension of the range of the species, but the phrase used is 
rather vague, as some of the headwaters of the Macquarie, which could be considered 
a "north-eastern tributary of the Darling", rise in the Blue Mountains, while other rivers 



120 



KOOT PAEASlTrSM IN ATKINSONIA T.IGUSTRINA, 



rising as far north as the Carnarvon Ranges in Queensland flow, occasionally at least, 
to the Darling. Specimens in the National Herbarium, Sydney, all come from the 
comparatively small area bounded by Linden, Mount Victoria and Mount Wilson, with 
one outlying locality at Marrangaroo. Material used in the present study came from 
Woodford and Linden. 

A. liyustrinu hap been reported (McKee. 1952) to form haustoria on the roots of 
several hosts, including Acacia intertexta Sieb., Platysace linearifolia (Cav. ) Norman 
(syn. Trachyniene linearis Spreng. ) and the narrow-leaved form of Leptospermuvi 
attenuatum Sm. Haustoria have also been observed on the roots of Monotoca acoparia 




Fig-. 1. — Cross-section of a root of Atkinsonia. 
Fig'. 2. — Young- roots of Atkinsonia (stippled) showing- branching and haustoria attacking 
tho host (solid blaclt). 

Fig-. 3. — An haustorium almost completely surrounding the host root. 



R.Br, and Cauntia sp. The haustoria examined in the present work were growing on 
the roots of Acacia intertexta, Leptospermum attenuatum and DiU-icynia ericifolia Sm. 
No pronounced differences in behaviour were observed between haustoria on any of 
these hosts. 

Root System and Haustoria. 
The late Dr. J. McLuckie made preliminary studies on the root system and 
haustoria of A. ligustrina. and raised seedlings which made considerable development 
in the absence of host plants. His seedlings, preparations and notes were unfortunately 
lost after his death in 1956. Seedlings were not studied during the present investigation, 
since a bush fire destroyed the above-ground parts of the plants from which our 
material was taken. The plants have since developed new shoots, but it will be some 



BY BARBARA P. MEA'zrp:S AND II. S. MCKEE. 



121 



time before they fruit again. For this reason the following account deals only with 
the root system and haustoria of established plants. 

A. ligustrina is a shrub with a large number of leading shoots, growing to a 
height of about one metre. The root system is entirely subterranean, consisting of Ave or 
six fleshy primary roots which rapidly turn blue when damaged. Lateral roots, bearing, 
the haustoria, are freely produced but short-lived. Thus the older primary roots are 
almost unbranched and bear numerous scars of fallen laterals. Young lateral roots are 
entirely devoid of root hairs, though Blakely (1922) recorded the presence of numerous 
root hairs on lateral roots of seedlings. 

Internally, the structure of the root is comparable with that of other dicotyledons. 
The apical meristera consists of a small group of cells giving rise to the root tissue 
and also to a small root cap of about three rows of cells. Differentiation of vascular 




Fig-. 4. — Cross-section through union of haustorium (stippled) and host (Leptospermum). 
The solid black represents vascvilar tissue of the haustorium and the broken lines, zones of 
collapsed cell.s. 



tissue tends to occur early, often less than lOO/i from the tip. The mature root is diareii 
with a well-developed endodermis with distinct casparian bands (Fig. 1). Secondary 
growth takes place in the normal manner. Periderm is superficial and limited in 
extent. Branch roots arise in the usual endogenous manner in the pericycle opposite 
a protoxylem group. 

Haustoria arise exogenously in the cortex of the root, their position being unrelated 
to the orientation of the parent root. They arise near the root tip, but their position 
is always lateral. They occur singly or in groups. It is possible that their development 
requires an external stimulus, perhaps proximity to a host root, but without seedlings 
this could not be investigated. The haustoria are often irregular in shape and may 
grow out some little distance before reaching a host root. The decayed remains of 



1'22 



ROOT PARASITISM Ii\ ATKriS'SONfA MGITSTRISA, 



numbers of young liaustoria were found: these had presumably failed to make contact 
with a host. 

When the host is reached, the haustorium flattens into a disc which attaches itself 
to the root (Fig. 2). This disc extends more or less evenly in all directions, wrapping 
itself round the host (Fig. 3). The disc may completely surround a small root, its 
originally separate parts .ioining with scarcely any line of division. 




Fig. 5. — Cross-section of parasite union showing elongated cells of the haustorium 
(.stippled) forcing apart the cells of the host xylem. The way in which the parenchyma ray 
cells have been pushed apart is particularlj' well shown. (Host Acacia.) 

Fig. 6. — Cross-section of union of parasite and host, showing the elongated cells at the 
margin of the haustorium (stippled) which have forced apart the xylem cells of the host 
(Acacia). 

Fig. 7. — Cross-section of part of an haustorium (cells left empty) growing through the 
region of the cambium of the host. In advance of the haustorium is a mass of yellowish 
material (dense stipijling), and some of the same material can be seen in the vessels in the 
adjacent xylem. Many partly dissolved cells can be seen in the cambial region. (Host 
Dillwyina.) 

Fig. 8. — Cross-section of host xylem and haustorium (stippled). In this older union the 
elongated cells at the margin of the haustorium are surrounded by yellowish material (shown 
black). Some elongated cells have entered the vessels of the host and grown along with them. 



The disc at this stage has a central axis of meristematic tissue surrounded by 
vacuolated parenchyma. The cells on the attaching surface are elongated with dense 
contents and large nuclei. At an early stage a zone of collapsed cells appears, extending 
from the axis to the edges of the disc. Similar zones are found in the haustoria of 
other semi-parasites. 



BY ]tAI!I)AI!A Jf. rMEN'/TES AXI) 11. S. MCKKK, 



123 



An outgrowth from the middle of the attached surface of the disc penetrates the 
host, partly dissolving, but chiefly forcing its way through the epidermis or cork into 
the cortex. Subsequent behaviour of the haustorium depends on the size of the host 
root. On a small root it may grow straight into the xylem, sometimes splitting the 
cylinder open. In a larger root, it may first extend tangentially in the host cortex, 
disrupting it to such an extent that groups of cortical cells may become embedded in 
the haustorial tissue. This process of radial growth and lateral expansion may be 
repeated in cortex, phloem, cambium and finally xylem, so that a complicated system ot 
overlying discs on a central shaft is produced (Fig. 4). Finally the haustorium may 
grow radially right through the middle of the xylem cylinder. 




Fig. 9. — I-X. Cross-sections cut at successive levels in the haustorium to show the 
changes in orientation of the vascular tissue. See te.\t for details. Phloem stippled, xylem 
clear, host xylem cross-hatched. 

Pig. 10. — Xylem union between host and haustorium : young stage in which cells of the 
haustorium (stippled) are closely adpressed to the vessels of the host. (Host Acacia.) 

B^ig. 11. — Xylem union between host and haustorium : older stage in which the haustorial 
cell in contact with the host vessel has been differentiated into a xylem vessel segment which 
links up with the main vascular tissue of the haustorium. (Host Acacia.) 



The advancing edges both of disc and shaft of the haustorium are composed of 
greatly elongated cells which have dense contents and large nuclei (Figs. 5, (5) and 
seem to force the opposing tissue apart. The cells are. nevertheless, quite plastic and 
in the xylem fit neatly into the gaps they have made (Fig. 5). In the phloem or cortex 
the cells reach their greatest size and are often embedded in masses of gummy 
material (Fig. 7). A similar gummy material is found in host xylem vessels near 
haustoria. It may result from a wound reaction by the host, or from the dissolving of 
host cells by the haustorium. Elongated haustorial cells sometimes grow inside vessels 
of the host. The margin of the haustorium is composed of glandular cells smaller 
than those just described, but also probably able to dissolve host tissue (Fig. S). 



124 HOOT PAHASITI8M IN ATKIXSONIA LKiTLSTUIN A, 

The development of the vascular system of the haustorium begins near the parent 
root with two or four groups of xj'lem, each with an associated phloem group (Fig. 9). 
Within a few millimetres these groups break up to fomn a small irregular ring of open, 
collateral bundles as in DendropMhoe falcata (Singh, 1954) and Loranthus niicranthus 
(Menzies, 1954). At first the bundles tend to lie side by side with the relative positions 
of the xylem and phloem alternating in alternate bundles. Then the number of bundles 
increases and they become arranged in an ellipse. Further down (but still outside the 
host) the proportion of vascular tissue increases, the xylem consisting of a solid, somc- 
v/hat squashed cylinder of cubical, reticulate vessels surrounded by what appears to be 
procambial or cambial tissue which differentiates into more xylem. No phloem was 
recognized at this level or below. Sections of the haustorium within the host plant 
show the xylem cylinder breaking up into a number of branches going out into the 
haustorial discs. 

Connections between xylem of host and parasite are easily found round the marginfi 
of the haustoria inside the xylem. In early stages a meristematic cell of the haustorium 




Fig. 12. — Section througii liost and para.site. No penetration of the host has yet taken 
place and a "gland" is a conspicuous feature of the haustorium. (Host Leptospermum.) 

is closely appressed to a host vessel (Fig. 10). This cell later matures into a reticulate 
or pitted vessel which is continuous with the vascular system of the haustorium 
(Fig. 11). Zones of collapsed cells occur in all parts of the haustorium, but the 
conspicuous zones found round each xylem connection in Lorarithus niicranthus are 
absent in Atkifisonia. 

The life of an haustorium appears to be relatively short. The haustoria found, 
varied from a millimetre to a centimetre in diameter, but this variation appears to be 
related to size of root attacked rather than to age. They were all yellowish in colour 
and in all the vascular tissue had developed to approximately the same extent Large 
woody haustoria with considerable secondary thickening, such as one sees in Loranthus. 
were not found. It seems probable that lateral roots and haustoi'ia last only about a 
year and are then replaced by new ones. 

Haustorial "Glaxds". 
In Hantalum alhtiiit, Cansjera rheeclii and Olax scanden.s ( Barbei', 1906) and also 
possibly in Thesium (Solms-Laubach, 1867-8), the haustorium appears to contain a 
"gland" secreting enzymes which function in the penetration of the host. A very 
similar structure has been found in the haustorium of Atkinsonia. The shape of this 
"gland" is shown in Figure 12. It first appears as two rows of elongated meristematic 
cells, those of one row abutting closely onto those of the other. As the "gland" becomes 



BY BAKBARA P. MKNZIES AXD H. S. MCKEE. 125 

older these cells become longer and longer and their contents begin to withdraw from 
contiguous walls. This continues until the "gland" consists of two rows of empty 
elongate cells surrounded by a very small-celled meristematic tissue (Fig. 13). Finally 
the elongated cells disappear altogether, leaving a cavity filled with red-staining 
amorphous material and fragments of cell and other debris. In Atkinsonia such 
"glands" were found only in haustorial discs which had not penetrated the host, but 
they were not found in all discs prior to penetration. No "glands" were found in older 
haustoria which had entered the host, nor was there any trace that a "gland" had ever 
been present. The "glands" appear to occupy the position taken up by the procambial 
tissue in haustoria which have penetrated the host. The orientation of the haustorial 
tissues round the "gland" also suggests that it is an abnormality following failure to 
penetrate the host rather than a transient feature of the normally developing 
haustorium. A possible sequence of events is that the penetration process grows out 




Fig. 13. — Detail of portion of Fig. 12 in the region of the "gland". The contents of the 
cells of the "gland" have contracted considerably. The tissues round the "gland" have dense 
contents and appear to have been considerably compressed. The density of the stippling 
indicates the density of the cell contents. 

from the middle of the attachment disc, but, unable for some reason to penetrate the 
host, is forced back into the attachment disc and its elongated cells begin to digest 
one another. This would account for the pieces of debris, apparently not belonging to 
the haustorium itself, which are found within old "glands". 

The structures figured by Barber (1906) for Santalum album show many 
similarities with those described above. Most of the "glands" he illustrates are in 
haustoria which have made little or no penetration, and he notes that the presence of 
a "gland" before penetration is not invariable. The peculiar structure he interprets as 
the "opening-up" of a "gland" would be more easily understood as a stage in "gland" 
formation on our interpretation. A similar example in a young haustorium of Atkinsonia 
has been found (Fig. 14): this could only be interpreted as the closing-up of a 
developing "gland". A difficulty encountered both with Santalum and Atkinsonia is 
that very few stages are found in which the haustorium is in the act of passing 



126 



KOOT PARASITISM IX ATKINSOXIA LIGTJSTKINA, 



through the cortex. This must be due to the fact that this stage is passed through 
very quickly. 

McLuckie studied the haustoria of Olax strieta R.Br., a brief summary of his results 
being given by McLuckie and McKee (1954), and some observations on the same 
species were made in connection with the present work. Olax also shows a haustorial 
gland somewhat resembling those found in Atkinsonia and Santaluvi. Further investi- 
gation on a wider range of material is desirable to permit detailed comparison of the 
structures in Olax with those of the other two genera. 




2.ooy- 



Fig'. 14. — Young- haustoriuni (stippled) in contact with tlie host. 

No reference has been found to a "gland" of this type in aerial members of the 
Loranthaceae. It is interesting that a structure found in two families of root parasites 
should occur also in a root parasite belonging to a family composed predominantly oL' 
branch parasites. 

Acknoicledgements. 
The authors wish to thank Miss I. Bowden, Woodford, N.S.W.. for indicating the 
position of the plants which provided most of the material studied; the late Dr. J. 
McLuckie for encouragement and co-operation in the early stages of the work; and the 
following for helpful correspondence or discussion: Dr. S. T. Blake (Brisbane), Dr. 
M. Cardenas (Cochabama, Bolivia), Prof. A. Guillaumin (Paris), Mr. L. A. S. Johnson 
(Sydney), Prof. D. Thoday (Bangor), Dr. R. Virot (Paris). 



ReferenQes. 
Anonymous, 1S69. — Memoir of W. H. Harvey, M.D., F.R.S. Bell and DaUly, London. 
Eentham, G., and Hooker, J. D., 1883. — Genera Plantarum, 3 : 212. Reeve, London. 
Barber, C. A., 1906. — Studies in root-parasitism. The haustorium of Santalum alum. 1. Early 
stages, up to penetration. Mem. Dep. Affric. in India, I (1) : 1-30. 



BY BAEBAKA P. MENZIES AND H. S. MCKEE. 127 

BlakelT;, W. F., is 22. — The Loranthaceae of Australia, Part ii. Proc. Linn, Soc, N,S.W., 

47: 199-222. 
Danser. B, H,, 1933, — A new system for the genera of Loranthaceae Loranthoideae, with a 

nomenclator for the Old World species of this subfamily, Verh. Akad. Wet. Avist. 2 Sect., 

29: 17. 
Danser, B. H., 1936. — The Loranthaceae Loranthoideae of the tropical archipelagos east of 

the Philippines, New Guinea and Australia, Bull. Jard. Bot. Buitenzorg 3 Sgr,, 14, 73-9S. 
EnglbRj a,, 1894. — Loranthaceae. In: Die Naturlichen Pflanzenfamilien. Auf. I, 3(1) : 156-198. 

Wilhelm Engelmann, Leipzig. 
Engler,, a., 1897, — Nachtrag zum II-IV TeAl Pflunzeiifamilien. Wilhelm EngelmoAvn, Leipzig, 
Engler, a,, and Krause., K,, 1935. — Loranthaceae, In : Die N atiirlichen Pflanzenfamilien. 

Auf, 2, 16b : 98-203, Wilhelm Engelmann, Leipzig, 
OuiLLAUMiN, A,, and Virot, R,, 1953, — Contribution & la flore de la Nouvelle-Caledonie, (Oil). 

Plantes recoltees par M. R. Virot. Hem. Mus. Nat. Hist. Nat. (Paris). Nouvelle Serio. 

Serie B, Botanique, IV, fasc. 1, ]-S2. 
Herbert, D. A,, 1918-19, — The Western Australian Christmas tree, Nuytsia florihundu, ils 

structure and parasitism, J. and Proc. Roy. Soc. W. Aiist., 5 : 72-88, 
HiERNj W. P., 1900, — Catalogue of the African j)la)its collected by Dr. Friederich Wehoitsch ijj 

1853-61, Dicots. Part IV, British Miiseiiui (Natural History), London, 
Hooker, J. D„ 1890, — Flora of British India. Vol. 5 : 207-8. 
Jackson, B. D., 1895. — Index Keivensis. 2 : 320. Ihiiversity Press, Oxford. 
LiNDLEY, J., 1839. — Sketch of the vegetation of the Swan River Colony. Appendix to tlii- 

index to the first 23 volumes of Edward's Botanical Register, London. 
MacBride, J. F., 1937. — Flora of Peru. Vol. XIII, Part 2, No. 2, Field Museum of Natural 

History, Chicago. 
McKee, H. S., 1952. — Root parasites in Loranthaceae. Natune, Lond,, 170: 40, 
McLucKiE, J,, and McKee. H. S., 1954. — Australian and New Zealand Botany. Associated 

General Publications, Sydney. 
Menzies, B. p., 1954. — Seedling development and haustorial system of Loranthus micranthns 

Hook. f. Phytomorphology, 4 : 397-409. 
Moore, C, and Betche, E., 19.^Z.~ Handbook of the Flora of Neio South Wales. Govei-ninent 

Printer, Sydney. 
Mueller, F. von, 1860. — Essay on the plants collected by Mr. Eugene Fitzalan during Lieut. 

Smith's expedition to the estuary of the Burdekin by Dr. Ferdinand Mueller, Government 

Botanist for the colony of Victoria. Melbourne. 
Mueller, F. von, 1861. — Fragm., 2: 130. 
Mueller, F. von, 1S65. — Fragm., 5 : 34. 

Oliver, D., 1880. — Hooker's Icones Plantarum. 3rd series, Vol. 4, Plate 1319. 
Rendlb, a. B., Baker, E. G., and Moore, S. Lh M., 1921. — A systematic account of the plants 

collected in New Caledonia and the Isle of Pines by Prof. R. H. Compton, M.A., in 1914. 

Part I. Flowering Plants (Angiosperms). J. Linn. Soc. (Bot.), 45: 245-417. 
Ruiz, H., 1940. — Travels of Ruiz, Pavon and Dombey in Pei'u and Chile ( im-lllS). Trans- 
lated by B. E. Dahlgren. Field Musewin of Natural History, Chicago. 
Singh, B., 1954. — Studies in the family Loranthaceae — list of new hosts of Dendrophthoe 

falcata (L.f. ) Ettings., its relations with hosts, the anatomy of its seedlings and mature 

haustorium. Agra Univ. J. Res., 3: 301-315, 
Skene, M,, 1924, — Biology of Floicering Plants. Sedgiuick and Jackson, London, 
Solms-Laubach, H, Graf zu, 1867-1868, — fiber den Bau und die Entwicklung der Ernahrungs- 

organe parasitisclier Phanerogamen, Pringsheim's Jb. wiss. Bot., 6 : 539-560, 
ViROT, R, 1956. — Le vegetation canaque. These, Faculte des Sciences de l'Universit6 de Pari.s. 



12!S 



DESCRIPTIONS OF TWO NEW SPECIES OF CURL'i AND ONE NEW SPECIES OF 

STIGMODERA ( BUPRESTIDAE ) . 
By C. M. Deuquet. 
(Three Text-figures.) 

[Read 29th April, 1959.] 



Synoiysis. 
Descriptions are given of two new species of Ctcris (C. aflamsi and C. zecki) from Central 
Queensland, and of one species of Stigmodera (.Stig. zecki) from Stanthorpe district, Southern 
Queensland. 



Cl'lfis ADAMSI, n. sp. 

Elongate, rather flat, especially the prothorax. 

Head bright green with flery reflections, clypeus brilliant green. Pronotum 
purple-blue on disc area with narrow medial vitta varicolorous, green being the 
dominant shade, side margins a fiery coppery colour. Elytra dark bluish, with the base, 
lower part of suture from past middle to 4 mm. from apex and mesosternal side bright 
metallic green showing flery coppery green reflections before their junction with the 
darker ground colour. Legs and antennae dark metallic blue. Underside: thoracic 
sternum fiery opalescent red, abdominal sternites purple. 

Head, fairly deeply impressed between the eyes. Pronotum much depressed, apex 
nearly straight, strongly bulging at the middle, base very irregularly sinuate, posterior 
angles acute and slightly produced, sides unusually wide behind middle; disc with a 
wide depression terminated by a large basal fovea; and two shallow lateral depressions 
finely punctate, the punctation more apparent at the brightly illuminated sides. 
8cutellum dark blue, subcircular, hardly depressed in centre. Elytra notably depressed, 
narrower at shoulder than the broad prothorax, almost parallel till past middle, 
widening slightly before the apical convergence; apices separately and neatly rounded, 
the hind margins strongly serrated, the elytra noticeably shorter than the abdomen. 
Four costae are clearly seen on each elytron, the intervals coarsely and irregularly 
punctate, the surface of the elytra being decidedly rugose. Underside finely and closely 
punctate, with a sparse white pubescence. Dimensions: 16 to 18 by 5 to 6 mm. 

Habitat: Edungalba, Central Queensland (E. E. .A.dams). 

A strikingly beautiful species of which four examples kindly sent to me by Mr. 
E Adams were examined. It is clearly distinct from previously described Guris, the 
nearest to it being Curis aurifera L. and G. and Guris olivacea Carter. It differs froin 
both (1) by the almost total absence of any bright coloration on the discal line of its 
pronotum and on its basal pronotal fovea; (2) by the greater length, depth and width 
of this central pronotal depression; (3) by the colour pattern of its suture; (4) by its 
more largely exposed pygidium. From Guris olivacea it, moreover, differs by the 
coarser texture of its elytra. 

Type in Coll. Adams; one paratype each to the Australian Museum, Sydney, and 
the author's collection. 

ClKIS ZECKI, n. sp. 

Oblong-ovate, somewhat attenuated at .apex. 

Head dark purplish blue, suffused with red gleams between the eyes; pronotum 
purplish blue with red reflections over the whole surface, but more so in the deep 
fovea at base; legs, antennae and underside dark metallic blue. Elytra purplish blue, 
the base, suture and sides bright green with golden red luminous radiations at their 
junction with the darker ground colour, this sutural fluorescence slightly interrupted 

Pkoceedings of the Linnean Society of New South Wales, 1959, Vol. Ixxxiv, Part 1. 



BY C. M. DEUQUET. 



129 



just past the base, widening a little past middle and coming to an end before the apex. 
Underside dark violet blue. 

Head widely but shallowly impressed between the eyes, closely punctate. Pronotuvi 
apex and base bisinuate, the former strongly produced at middle, anterior angles acute, 
posterior sharply produced; disc with medial sulcus almost throughout, terminating 
with a deep and wide fovea at middle near base; finely and regularly punctate. 
Scutellum roundish, glossy, impunctate. Elytra wider than prothorax, shorter than 
body, slightly enlarged at shoulders, subparallel to past middle, thence tapering to 
apex; apices widely and separately rounded, margins neatly denticulate on apical part, 
disc irregularly punctate with three smooth costae, a considerable area of pygidium 




exposed. Underside finely punctate, with a short cinereous pubescence. Dimensions: 
15 X 5 mm. 

Habitat: Edungalba, Central Queensland (E. E. Adams). 

Type in Coll. Adams. One paratype each to the Australian Museum, Sydney, and 
to Mr. Zeck's and the author's collections. 

This fine and clearly distinct species was discovered by Mr. E. Adams, an excellent 
and active entomologist from Central Queensland. Six examples are before me for 
examination. While showing some similarity to Curis yalgoensis Carter, it differs 
from it in (1) slightly smaller size, (2) colour of pronotum and elytra which is more 
florid in yaJgoerisis, (3) wider and deeper depression near base of pronotum, (4) 
different width of sutural flery coloration, (5) wider apex, (6) pygidium more largely 
exposed. 

In happy memory of a 40 years' sincere friendship I have given Emil Zeck's name 
to this rare insect. 

Stigmodeea zecki, n. sp. 

Ovata forma, paulum ad apices attenuata. Capite thoraceque aereo-viridibus, 
nitidis et punctulatis; elytris bronzeo-nigris, nitendibus, circum suturam majore aereis; 
maculis ternis flavis, apicis marginibus externis splendide sanguineis, punctato-striatis, 
pedibus viridi-aeneis; corpus subtus viridi-aeneum, dense albo pubescens. 

Ovate, slightly attenuated behind. 

Head fairly bright bronzy green; legs metallic bluish, antennae green; pronotum 
green with coppery gleams; elytra glossy dark blue with greenish reflections tinged 
I 



130 



NEW SPECIES OF CURIS AND STIGMODEEA. 



with bronze in the sutural area, with pale yellow markings as follows: female with 
two small roundish spots near the base almost above the shoulders, two slightly larger 
squarish ante-median ones situated behind the former, neither of these spots extending 
to the sides, a very small lateral one on the exterior margin situated about one-third of 
the entire length of the elytra from their base between the two previous spots and a 
straight anteapical fascia extending to the sides, but interrupted at suture and produced 
backwards along the margins down to within two mm. from the apex, this latero- 
produeed part brightly sanguineous. With the male, the two roundish spots near 
the base are replaced by an oblique vitta produced downwards to meet and include the 
two medial spots. 

Head excavated between the eyes, very closely punctate. Pronotum almost straight 
in front, strongly bisinuate at base, all angles acute, the posterior ones strongly 
produced, sides arcuately enlarged from apex to base, disc without medial line but 
showing a small basal fovea in front of the scutellum; fairly deep impressions at sides. 
Scutellum. subcordate, centrally impressed. Elytra sharply widened behind shoulders 
and also past middle, fully attenuate behind, striate-punctate, the intervals more 
strongly raised on apical half, rounded and only minutely punctured; apex widely 
lunate, each elytron with wide oblique lunation and bispinose, the external spine more 
prominent, the sutural minute. Underside bronzy green, sternum with coppery tint, 
abdomen finely punctate, clothed with recumbent cinereous hair. Dimensions: 13 x 4i 
mm. 

Habitat: Stanthorpe district. Southern Queensland (A. Gemmell and E. Sutton). 

Type in the Australian Museum, Sydney. Paratypes in E. Zeck, A. Gemmell, E. 
Sutton and the author's collections. 

This pretty Buprest belongs to the interesting producta group which is confined 
to the eastern States of Australia, although the habitat of Stigmodera producta itself 
when described by Edw. Saunders 90 years ago was inaccurately given as Western 
Australia. Five examples are before me for examination. It is closely allied to Stig. 
delta Thomson, but differs as follows: 



Stig. delta Thorns. 
Average size: 14 x 5 mm. 
Form more elongate. 
Colour duller, less nitid. 
Pattern: The yellow colour of the basal 

vitta of the elytra nearly always 

reaches the margins. 
Thin and smooth pronotal median line 

clearly visible. 
Interstices more raised. 

Marginal yellow stripe between the second 
and third leg wider and longer. 

Sides greatly attenuated to the apex 
which terminates with minute oblique 
lunation. 



Stig. zecki, n. sp. 
13 X 4J mm. 
Form more rotund. 
More glossy, more glittering. 
It never does reach them, neither with 
males nor females. 

Absence of pronotal median line. 

Interstices less prominent, surface of 

elytra smoother. 
Marginal stripe thinner and shorter, 

almost invisible in some examples. 
Sides widely lunate with much longer 

external spine. 



It may be mentioned that the yellow preapical fasciae and the dark blue preapical 
patches of S. delta and 8. zecki are very different in shape and colour, and the back- 
wards extension of the sanguineous margin also shows dissimilarities in width and 
length between these two species. Also that the ornamental design and coloration of 
the elytra of 8. zecki is an almost exact replica of that of the dark variant of 8. 
delectabilis Hope found in the Illawarra Coastal Range of N.S.W. 

As a mark of appreciation of Mr. Emil Zeck's most valuable contribution to 
Australian entomology, it gives me great pleasure to name this new species after him. 



131 



NOTES ON AUSTRALIAN MOSQUITOES (DIPTERA, CULICIDAE). IV. 

Aedes alboannulatus Complex in Victoria. 
By N. V. DoBROTWOKSKY, Zoology Department, University of Melbourne.* 

(Four Text-figures.) 
[Read 29th April, 1958.] ' ■ 



Synopsis. 
The present study has revealed that Ae. alhoannulatus complex in Victoria consists of 
six forms. A comparative account of the biology and ecology of Ae. queen^landis and of the 
new forms is given. The taxonomic status of three forms is discussed. Adults of both sexes 
and the larvae of Ae. rupestris, n. sp., and Ae. titbbutiensis , n. sp., are described. An account 
is given of the variability of Ae. queenslaouHs (Strickl.) and notes on Ae. alboannulatus 
(Macq. ) are added. 



Most species of the alboannulatus group are distributed in the Oriental region and 
in the northern part of the Australian region. Only some species of alboannulatus s. str. 
subgroup (Knight and Marks, 1952) extend south into Victoria, South Australia and 
Tasmania. The two most common species of this subgroup, Ae. alboannulatus and 
Ae. queenslandis, have been previously recorded from Victoria (Edwards, 1924). 

Ae. alboannulatus varies very little over its range of distribution and only one 
conspicuous variation, a yellowness of the apical part of the femora, has been found; 
this variety occurs in the western part of the Otway Ranges. 

Ae. queenslandis, on the other hand, has been regarded as a highly variable species. 
A detailed study of specimens from Victoria has confirmed this opinion, but has also 
shown that the name Ae. queenslandis has, in the past, been applied to a complex of 
closely related forms. In addition to Ae. queenslandis four others can be recognized. 
Three of them have different geographical and/or ecological distribution and do not 
interbreed where their distribution overlap. Two are treated as species and are 
described below under the names Aedes rupestris and Aedes tubbutiensis ; the third 
(Form A), which is undoubtedly a distinct species, will not be discussed here since it 
has already been reviewed, though not as yet named, by Lee, Dyce and O'Gower (1957). 
The fifth member of the complex, which will be referred to as the "yellow form", 
appears not to be reproductively isolated from queenslandis and will be treated as a 
variety of this species. 

Key to Victorian species of alboannulatus complex. 
Adults. 

1. Tibiae and proboscis mottled with pale scales 2. 

Hind tibiae and proboscis black 3. 

2. Scutum with patches of white scales. Femora v/ith ochreous preapical ring, which may 
extend apically. Posterior pronotum with bronzy narrow scales dorsally. Venter' white 
with median black spots alboannulatus. 

- No patches of white scales on scutum. No preapical ring on femora. Posterior pronotum 
with pale goldish narrow scales dorsally tubbutiensi'S, n. sp. 

3. Tibiae with white sub-basal ring form A. 

- Tibiae without sub-basal ring 4. 

4. Venter ochreous usually with scattered black scales queenslandis. 

- Venter black with white lateral patches. Apical border of sternites always with broad 
black band rupestris, n. sp. 

* This work was supported in part by a grant from the Trustees of the Science and 
Industry Endowment Fund of Commonwealth Scientific and Industrial Research Organization. 

Proceedings of the Linnean Society of New South Wales, 1959, Vol. Ixxxiv, Part 1. 



132 , NOTES ON AUSTRALIAN MOSQUITOES. IV, 

Larvae. 

1. Head setae 4, 5 and 6, almost in straight line. Prothoracic seta 1, 2-branched ; 4, single : 
5, 2-3-branched rupestris. 

- Head setae 4, 5 and 6, forming triangle. Prothoracic seta 5, single 2. 

2. Head setae 4, 5 and 6, forming almost a right angle, with setae 4 well in front of setae 5. 
Prothoracic seta 1, 2-branched ; 4, single alboannulatus. 

- Head setae 4, 5 and 6, forming obtuse triangle with setae 4 slightly in front of or behind 
setae 5 3. 

3. Prothoracic seta 1, 2-branched ; 4, 2-branched, rarely single 

queenslandis. 

- Prothoracic seta 1, single, rarely 2-branched on one side; 4, 2-3-branched 

tubbutiensis. 

Distinguishing the larvae of rujiestris, alboannulatus and queenslandis is not 
difficult, but the larvae of tubiutiensis are very similar morphologically to those of 
queenslandis and some cannot be identified with certainty. 

Aedes ALBOANNULATUS (MacQuart). 

Gulex albo-annulatus Macquart, 1849, Dipt. Exot., suppl. 4: 10. 

Distinctive Characters: Adult. Vertex clothed with narrow curved bronze scales 
becoming white towards centre of vertex. Proboscis mottled. Scutum clothed with 
narrow, curved bronze scales and some patches of white scales. Posterior pronotum 
with narrow curved bronze scales dorsally. Femora with preapical band; femora and 
tibiae mottled. Sternites white scaled with median patch of black scales. 

Description of Adult. 

Female. — Head: Vertex clothed with narrow curved bronze scales except for small 
median area of white scales. Upright scales black; lateral scales broad and white 
except for large black patch in middle. Palpi black scaled with white scales on base 
of segments and on apex of last. Proboscis black scaled, mottled or pale scaled m 
middle. Scutum clothed with narrow curved bronze scales with some small areas of 
white scales; a patch of broad white scales just in front of scutellum. Anterior 
pronotum with narrow curved and elongate white scales and black bristles. Posterior 
pronotum with broad white scales below, broad black scales in middle and narrow 
bronze scales above; black area may have a few pale scales. Tergites black with 
Incomplete white basal bands and lateral spots. Sternites white scaled with median 
and lateral apical black spots. Wing length: 5-0-7-5 mm. Upper fork cell 1-5-2-3 times 
as long as its stem. Legs: femora with preapical band or patch, which may join with 
knee spot. Tibiae mottled. Fore and mid tarsi with 2-3 white basal rings, hind tarsi 
with 4. 

Male. — Palpi about as long as proboscis without labella; last two segments, and 
apex of shaft, with long, dark, silky hairs; a white spot at base of each segment. 
Proboscis less mottled than in female, sometimes only a few white scales. Upper fork 
cell 1-3 times as long as its stem. Tergites 2-7 with complete white basal bands; 2 may 
have only a patch of white scales; tergite 8 may be mottled. Sternites usually black 
with lateral white spots; white scales may predominate and black scales be reduced 
to patches in middle of segment and apical corners. Terminalia: very similar to 
terminalia of queenslandis and can be recognized only by the basal lobe of the coxite 
(Fig. 1, a) which has a row of about 20 setae instead of 10 as in queenslandis. Appendage 
of harpago not as expanded in middle as it is in queenslandis. Ninth tergite with 
prominent lobe bearing 2-5 strong setae. 

Larva. — Head and siphon brown. Head about five-sevenths as long as broad. Antenna 
about half length of head, clothed with spicules; seta 1, 3-5-branched, arising at about 
half-way from base. Head seta 4, 4-5-branched ; 5, 3-5-branched; 6, 2-4-branched; 7, 6-7- 
branched; 8, 1-2-branched; 9, 2-3-branched. Head setae 4, 5 and 6 arranged to form 
apices of almost right angle triangle (Fig. 1, d, c). Seta 4 medial to 5 and 6 and 
between them. Mentum with 9-12 lateral teeth. Prothoracic setae (Fig.l, &) : 1, 2-branched; 
2, single, only slightly shorter than 1; 3, 4-5-branched, about three-quarters length of 2; 



BY N. V. DOBROTWOKSKY. 



133 



4, single, about as long as 3; 5 and 6, single; 7, 3-branched. Eighth abdominal segment: 
Pentad setae: 1, 3-5-branched ; 3, 8-11-branched; 5, 4-6-branched; 2 and 4, single. Comb 
of large patch of fringed scales. Siphon index: 2-6-3-4, mean 2-9; seta 1, 6-7 -branched. 
Pecten of 17-20 spines, each with 3-4 teeth at base, central ones the largest. Anal 
segment: saddle covering about half segment. Seta 1, single or 2-branched; 2, 5-8- 
branched; 3, single; 4, of 15 tufts. Anal papillae about as long as saddle, or shorter. 
Eggs (Fig. 1, g) black, narrow, slightly flattened ventrally, about l-O-l-l m. long with 
an index of about 4. 

Biology: Ae. alboannulatus is a sylvan species. It occurs almost throughout 
Victoria, but has not been recorded in the Mallee. It breeds in ground pools and rock 
pools; the water is usually clear but may be more or less cloudy. Ae. alhoannulatus 
usually avoids heavily shaded pools in dense forest, showing a preference for diffuse 






6,9 p.. 






Fig. 1. — Aedes alboanmilatus (Mocq. ). Adult: a. basal lobe of coxite. Larva: b, pro- 
thoracic setae ; c and d, bases of head setae 4, 5 and 6 ; e^ pecten tooth ; f, mentum ; g, egg, 
plan and side view. 

sunlight. On the Bogong High Plains, at an altitude of 5,400 ft., it was found during 
the summer to be breeding in ground pools fully exposed to the sun. In the flat country 
north of the Dividing Range it can use such pools only during the cooler months of 
the ysar, being confined during the summer to shaded gullies, backwaters and roadside 
ditches. 

The spring generation of albonnnulatus oviposits on the edges of pools many of 
which will dry during the summer; the eggs then remain dormant until autumn rains 
fill the pools again when most of the eggs hatch within 12 hours. Some larvae pupate 
and produce adults, early in August, but the majority do not pupate until the spring. 
In permanent water pools alboannulatus is able to breed all the year round. 

Adults are always abundant during the spring, and commonly during the autumn 
in favourable places, but a prolonged dry summer may result in the drying out of 
pools which in normal years permit continuous breeding. 

In Victoria alboannulatus is sometimes found breeding alone, but usually it is 
associated with Ae. queenslandis, G. fergusoni Taylor, C. pipiens australicus Dobr. and 
Drumm., Th. inconspicua Lee and A. annuUpes Walk. 



134 NOTES ON AUSTRALIAN MOSQUITOES. IV, 

Habits: The adults are vicious day-biting mosquitoes; they bite even during the 
T/inter at temperatures as low as 11°C. 

AiiDES QUEENSLANDis (Strickland). 

Culicelsa queeyislandis Strickland, 1911, Entomologist, 44, 179. Culicelsa similis 
Strickland, 1911, ibid., 44, 132. Culicada demansis Strickland, 1911, ibid., 44, 202. 
Culicada cmnpstoni Taylor, 1914. Trans. Ent. Soc. Lond., 1913, 692. Culicada hybrida 
Taylor, 1916. Proc. Linn. Soc. N.S.W., 41, 568. 

This is a rather variable species, and hence it is difficult to specify morphological 
traits which are common to all variants. 

Distinctive Characters: Adult, Female. — Vertex with large or small area clothed 
with golden-yellow scales. Proboscis black scaled. Integument from light brown to 
dark brown. Thorax clothed with narrow light yellow or bronze scales; no patches of 
broad white scales in front of scutellum. Posterior pronotum with patch of broad 
white scales below, broad black scales in middle and goldish, narrow, curved scales 
above. Tarsi of fore and mid legs with two basal white bands, hind legs with four. 
Tergites black with incomplete yellowish basal bands, sometimes reduced to a few pale 
scales. Sternites clothed with ochreous scales; lateral apical spots of black scales, or 
of black scales and ochreous mixed. 

Male. — Sternites black scaled with lateral elongate white spots; black apical part 
of sternites with some ochreous scales. 

Larva. — Head setae 4, 5 and 6 arranged to form apices of a triangle; setae 4 lie 
between setae 5 and slightly in front of or slightly behind a line drawn through them. 
Prothoracic seta 1, 2-branched; 4, 2-branched or single, shorter than 3; 5, single. 

Description of Adult. 

Female. — Head: Vertex usually with large patch of golden yellow scales. In 
Tasmanian specimens this patch greatly reduced and all upright scales black. Proboscis 
black scaled. Thorax: Integument usually light brown, but dark brown in Tasmanian 
specimens. Scutum clothed with narrow curved golden yellow scales. Laterally, 
particularly in area near scutal angle, there is admixture of dark bronze scales. 
Scutum of Queensland specimens clothed with almost uniform goldish yellow scales; 
in Tasmanian specimens it is mainly clothed with dark bronze, almost black, scales, 
with yellowish bordering scales. Prescutellar area without broad scales. Anterior 
pronotum with yellowish narrow curved and elongate scales and bristles; in some 
specimens, particularly Tasmanian, bristles on dorsal part of anterior pronotum are 
dark, almost black. Posterior pronotum with patch of broad white scales below 
(yellowish in some Queensland specimens), broad black scales in middle and goldish 
yellow narrow curved scales above. South Australian and Tasmanian specimens have 
some pale broad scales in black middle area. Area of narrow yellowish scales reduced 
in some specimens to a line, but in others it extends downwards, reducing the black 
area. Pleura with usual patches of broad creamy scales and bristles. Wing length 
3-2-5-4 mm. Upper fork-cell 2-3-3 times as long as its stem. Fore and mid femora 
black with mottling of yellowish scales, except for ventral side of basal two-thirds, 
which is pale scaled. Hind femur pale on basal half, black with yellowish mottling on 
apical half. In some specimens mottling increased towards end of femur and forms 
small or large ochreous preapical patch or ring; in some specimens apical part of 
femur is yellow. Tibiae black, pale scaled posteriorly, sometimes with streak anteriorly. 
In a few specimens hind tibia with inconspicuous or incomplete white ring at base. 
First 2 or 3 tarsal segments of fore and mid legs with white bands; tarsi of hind leg 
with 4 bands, some specimens with narrow fifth band. Band on segment 4 usually half 
length of segment, but in Queensland specimens it is only one-quarter-one-fifth of length 
of segment. Abdomen: Tergites black scaled with white lateral spots and incompletf; 
yellowish basal band on tergites 2-6; in some specimens bands reduced to a few pale 
scales. Tergites 7-8 black or mottled with yellowish scales; in some specimens yellow 



BY N. V. DOBEOTWORSKY. 



135 



scales increased on tergites 6-8 replacing black scales. Sternites clothed with ochreous 
scales, with apical lateral black spots, and usually mottling of black scales, the number 
of which sometimes exceeds yellow ones. 

Male. — The male differs from the female as follows: Palpi black scaled, as long as 
proboscis without labella. Last two segments and apex of shaft with long dark hairs; 
segments 2-5 with patch of white scales. Torus brownish; flagellar segments pale, 
with dull silky verticillate hairs. Vertex clothed with narrow curved and upright 
yellowish scales. Thorax: Integument brown. Scutum clothed with yellow goldish 
scales with some admixture of dark bronze scales near area of scutal angle. Femur 
of fore leg black, mottled laterally and ventrally; femur of mid leg black mottled, 
pale scaled ventrally on basal half; femur of hind leg pale on basal two-thirds, apically 
black with few yellowish scales. Fore tarsi with 2 white rings, mid tarsi with 3, hind 



4 S 




96 



,05 



a;:i 




Pig. 2. — Aedes queenslandis (Strickl.). Adult: a. basal lobe of coxite. Larva: b, prothoracie 
setae ; bases of head setae 4, 5 and 6 from : c, Queensland ; d, Victoria ; e, Tasmania ; f, 
mentum ; g, pecten tooth ; h, egg, plan and side view. 



tarsi with 4. Ring of segment 1 about one-fifth of segment. Tergites 2-6 with white basal 
bands, usually complete. Sternites black scaled with elongate lateral white patches not 
reaching apical border, and some yellowish scales scattered apically in black area. 
Terminalia: Coxite clothed basally with white scales and apically with black scales 
and golden bristles. Style about half length of coxite, narrowing to both ends, with 1-2 
preapical setae, appendage slightly curved and about half length of style. Basal lobe of 
coxite narrow, transverse, with row of about 10 setae (Fig. 2, a). Harpago stout, with 
fine setae at base; appendage about as long as harpago, widened at middle and narrowiiig 
to end. Paraproct with single tooth and 5 fine setae. Ninth tergite with prominent 
lobe bearing 2-7 stout setae. 

Variability of Males. — Scutal scales may be dark or light golden, with some 
admixture of dark bronze scales. In some specimens black scales occupy the greater 
part of the posterior pronotum. Upper fork-cell 1-4-1-7 times as long as stem. Sternites 
sometimes clothed with mixture of black, yellow and white scales in different pro- 
portions, or may have black median and lateral apical spots and white lateral spots. 



136 notes on attstralian mosquitoes. iv, 

Yellow Form. 

This form differs from other variations in its general yellow colour, but, although 
typical specimens are quite distinct from the type form of queenslandis, there are 
intermediates. 

The yellow form is characterized by: Lighter integument. Proboscis with a few 
pale scales on basal half. Torus yellow, darker on inner side. Broad scales on lower 
part of posterior pronotum yellowish. Fore and mid femora intensively mottled with 
yellowish scales, hind femora yellowish with mottling of black scales. Fore tibia pale 
below, black above — mid tibia mottled with yellowish scales; hind tibia black abova 
and apically, elsewhere yellow. Tarsi of fore and mid legs with 3 banded segments, 
hind with 4; band on segment 4 about one-fourth length of segment. Bands on tergites 
in females, reduced to yellow patches or absent. Last 2-3 tergites clothed with ochreous 
scales. Sternites clothed with ochreous scales and a few scattered black ones; apical 
lateral spots black. 

Larva. — Head and siphon light brown. Head about three-fourths as long as broad. 
Antenna about half length of head, clothed with spicules; seta 1, 4-6-branched, about 
half length of antenna and arising about halfway from base. Head seta 4, tiny, 4-6- 
branched; 5, 4-9-branched ; 6, 3-5-branched; 7, 8-10-branched; 8, single; 9, 2-3-branched. 
Setae 4, 5 and 6 arranged to form apices of triangle; setae 4 lie between setae 5, slightly 
in front of, or slightly behind, a line drawn through them (Fig. 2, c, d, e). Mentum 
with 12-13 lateral teeth. Prothoracic setae (Fig. 2, b): 1, 2-branched; 2, single; 3, 5-10- 
branched, about three-fifths as long as 2; 4, small, 2-branched or single; 5 and 6, single; 
7, 3-branched. Eighth abdominal segment; Pentad setae; 1, 4-6-branched; 3, 8-13- 
branched; 5, 4-9-branched; 2 and 4, single. Comb, large patch of fringed scales. 
Siphon index 3-1-3-7; seta 1, 8-9-branched. Pecten of 19-26 spines, each with 4-5 teeth 
at base. Anal segment: saddle covering about half segment. Seta 1, 1-2-branched; 2, 
6-11-branched; 3, single, long; 4, of 13-16 tufts. Anal papillae about as long as saddle. 

Eggs (Fig. 2, h) black, oval, slightly flattened ventrally and narrowing to the 
posterior end. They are about l-O-ll mm. long with an index of 3-2-3-6. 

Biology. — Ae. queenslandis is confined to woodlands. It breeds in swamps, ground 
and rock pools, dams, cavities in logs, and in artificial containers (tanks, tins, etc.), 
usually in shaded situations, since it requires water of relatively low temperature. It 
may make use of exposed shallow pools during the cooler months of the year, but in 
the summer can do so only at high altitudes; deep exposed pools are always more 
suitable. 

In Victoria the temperature of the water in which Ae. queenslandis breeds has not 
been found to exceed 18°C., even during the summer, but in southern Queensland, where 
again shaded pools are preferred, the larvae may be able to tolerate higher tempera- 
tures; they have been collected (I. C. Yeo and H. Grening, 6.2.58) in pools exposed to the 
sun for several hours in which temperatures by mid-morning had reached 20 °C. 

The water may be clean or contain decaying leaves. In Victoria queenslandis some- 
times breeds alone, but usually is found in association with Th. ineonspicua Lee, C. 
ferffusoni Tayl., Ae. alboannulatus Macq., and more rarely with Th. littleri Tayl., A. 
stigmaticus Skuse, A. pseudostigmaticus Dobr., Ae. notoscriptus Skuse, Ae. rupestris 
Dobr., T. tasmaniensis Strick., and with most of spring breeding Aedes of the subgenus 
Ochlerotatus. During the coldest months development of larvae is slow, but a few 
pupae can be found almost throughout the winter. In the laboratory emergence of 
adults was observed at 10-11°C. Ae. queenslandis oviposits on moist soil or rock 
surfaces just above water level. The eggs cannot withstand severe desiccation, but in 
natural conditions, in forests, they remain viable throughout the summer and hatch 
after autumn rains fill the pools. 

AiiUES RUPESTRIS, U. Sp. 

Types. — The type series were bred from larvae collected at Lome. All specimens 
have their associated larval and pupal skins. The holotype female, allotype male, six 
paratype males and six paratype females are in the collections of the National Museum, 



BY N. V. DOBROTWORSKY. 137 

Melbourne. One paratype male and one paratype female are in each of the following 
collections: C.S.I.R.O., Division of Entomology, Canberra; School of Public Health and 
Tropical Medicine. Sydney; University of Queensland, Brisbane; British Museum 
(Natural History), London; U.S. National Museum, Washington. 

Distinctive Characters: Adults.^ — Narrow curved scales on vertex goldish. Integu- 
ment dark brown. Scutum with narrow dark golden and small areas of pale scales 
laterally. Scales in front of scutellum narrow. Femora mottled anteriorly. Proboscis 
and tibiae entirely black. Sternites black with white lateral patches. Apical border 
always with broad black band. Ninth tergite of male without prominent lobes. Larva: 
Head dark brown. Head setae 4, 5 and 6 with their bases about on a straight line. 
First and fifth prothoracic setae, 2-branched. 

Holotype Female. — Head: Vertex with narrow curved decumbent scales, yellow 
goldish in centre. Upright forked scales yellow golden, becoming dark towards sides 
and neck. Lateral scales broad, flat and white except for central black patch. Torus 
black with patch of white scales; first flagellar segment of antenna pale at base with 
patch of white scales; other segments black. Palpi black scaled with patch of white 
scales on segment 2 and at base and apex of segment 3. Thorax: Integument dark 
brown. Scutum clothed with narrow dark golden scales, becoming pale around bare 
area, in front of wing roots, along border with posterior pronotum and anteriorly; 
admixture of black scales laterally. Scutellum pale scaled. Anterior pronotum with 
dark bristles, goldish curved narrow scales and broader elongate scales. Posterior 
pronotum with patches of broad white scales below, broad black scales in middle and 
narrow curved goldish scales above. Pleura with usual patches of broad white scales 
and pale bristles. Wing length: 4-2 mm. Upper fork-cell about twice the length of its 
stem. Stem of halteres pale; knob pale scaled except at tip and underneath, which are 
black scaled. Legs: Fore and mid femora mottled, pale ventrally on basal half; tarsal 
segments 1 and 2 with white bands, 3 with a few white scales at base. Hind femur 
pale ventrally on basal two-thirds, apical third black scaled with few white scales 
posteriorly. Hind tarsal segments banded, except last one. Band on segment 4 half 
length of segment. All legs with creamy knee spot and black tibia. Abdomen: 
Tergites black scaled. First tergite with apical patch of black scales; 2-5 with incom- 
plete narrow white basal bands; 6 and 7 each with patch of pale scales. White lateral 
spots on all segments. Sternites black scaled with white lateral patches. Apical border 
of sternites always with complete broad black band. 

Paratype Females. — The series of 11 paratype females does not show much varia- 
tion. Length of wing varies from 3-6 to 4-3 mm. In some specimens tergite 1 has an 
apical patch of black scales mixed with a few white, 2 has a basal patch of white scales; 
3-7 have complete, narrow bands. Sternites may be pale scaled with inconspicuous 
black scales, but apical black band is always conspicuous and complete. 

Allotype Male. — This differs from the holotype as follows: Palpi black scaled, as 
long as proboscis without labella. Last two segments and apex of shaft with long dark 
hairs; 2-5 segments with patch of white scales. Torus black, flagellar segments pale, 
with dark silky verticillate hairs. Tarsi of fore and mid legs with 3 white basal bands, 
hind tarsi with 4. Wing length: 3-9 mm. Upper fork-cell 1-1 times length of stem. 
Abdomen: Tergite 1 with few dark and pale scales in middle, 2 with an incomplete 
basal band, 3-6 with basal bands joining white lateral spots, 7 with a few white scales 
at base, 8 with lateral white spots only. Sternites black with white elongate lateral 
spots not reaching apical edge. Terminalia: Coxite black scaled, with patch of white 
scales basally; laterally and sternally it bears long and some short setae. Style half 
length of coxite, narrowing sharply at mid length; 1-2 preapical setae; terminal 
appendage straight, about half length at style. Basal lobe of coxite narrow, transverse, 
with row of about 10 long setae along edge and several small ones on upper side. 
Harpago stout, with fine setae at base; appendage about as long as harpago, widened at 
middle and narrowing to end. Paraproct with single tooth and 2 very fine setae near 
tip. Ninth tergite without prominent lobe, a few fine short setae on inner side. 



138 



NOTES ON AUSTRALIAN MOSQUITOES. IV, 




Pig-. 3. — Aedes rupestris, n. sp. Adult: a, male termlnalia ; b, basal lobe of coxite; c, 
harpago. Larva : d, prothoraclc setae ; g, head, terminal seg-ments and mentum. Pupa : e, part 
of cephalothorax ; /, dorsal view of terminal abdominal segments ; h, egg, plan and side view. 



BY N. V. DOBEOTWORSKY. 139 

Paratype Males. — The series of 11 paratype males does not show much variation. 
The palpi of some specimens are slightly shorter than the proboscis. Some have only 
2 tarsal bands on fore legs. Wing length: 3-6-3-9 mm. Sometimes tergite 7 has a 
complete basal band. White lateral spots on sternites are reduced in some specimens. 

Larva (Fig. 3, d, g). — Head and siphon dark brown, body blackish. Head about 
four-fifths as long as broad. Antenna about two-fifths length of head, clothed with 
spicules, seta 1 with 2-4 branches, less than half length of antenna and arising at 
about two-fifths length from base. Head seta 4, tiny, 4-7 -branched; 5, 5-7-branched; 6, 
3-5-branched; 7, 5-10-branched; 8, single; 9, 2-3-branched. Setae 4, 5 and 6 with bases 
almost in a straight line. Mentum with 8-10 lateral teeth on each side. Prothoracic 
chaetotaxy: seta 1, 2-branched, long; 2, single, about three-fifths length of seta 1; 3, 
stellate, 5-11-branched and about half length of seta 1; 4, short, 1-2-branched; 5, long, 
2-3-branched; 6, single, as long as 5; 7, 3-branched. Eighth abdominal segment: Pentad 
setae: 1, 3-4-branched; 3, 7-13-branched; 5, 5-7-branched; 2 and 4, single. Comb of 
about 150 fringed scales. Siphon stout, index 2-6-3-5, mean 2-9, seta 1, 6-10-branched. 
Pecten of 16-28 spines each with 2-3 teeth at base, the upper the largest. Anal segment: 
saddle covering more than half the segment. Seta 1, usually 2-branched, sometimes 
single; 2, 6-10-branched; 3, single, long; 4 of 14 tufts (one or two precratal). Anal 
papillae three-fifths to four-fifths length of saddle. 

Eggs (Fig. 3, h) black, almost regular oval, flattened slightly on ventral side. 
They are about 1-0 mm. long with index 2-9-3-2. 

Biology. — The breeding places of Ae. rupestris in Victoria are rock pools, in 
eucalypt forests, which are exposed to the sun. These are usually small and shallow 
with a thick layer of black mud and decayed leaves on the bottom. The water is brown, 
with a strong smell of decomposed eucalypt leaves. The temperature of the water in 
such pools rises during summer to about 30°C. at 1-2 p.m. In south Queensland the 
larvae of this species were collected, in large numbers, in two rock pools exposed to 
the sun; at 1-2 p.m. the temperature of water exceeded 37-5°C. (E. N. Marks, 19.11.57). 

In some rock pools, particularly shallow ones with sloping edges, Ae. rupestris 
alone was found during the summer, but in these same pools, from late autumn to early 
spring, there were always some queenslandis larvae. In large and deep pools, with 
cleaner and cooler water, or with some vegetation, queenslandis was always more 
numerous than rupestris even during the summer. Ae. alboannulatus was found on 
several occasions in association with rupestris. 

In south Queensland it was found in association also with A. annulipes Walk., 
C. halifaxi Theob. and C.j). australicus Dobr. and Drumm. 

The eggs which remain dormant on the edges of dried pools hatch after being 
submerged by later rains. In the laboratory hatching of eggs was observed at tempera- 
tures in the range 9-14-5°C. At such low temperatures larvae developed only slowly, 
taut eventually produced adults. Successful emergence of adults was observed in the 
laboratory at temperatures as low as 10°C. 

Hahits. — It is a day-biting mosquito, and is very common near its breeding sites; 
it ceases biting early in the winter when the temperature falls to 16-17°C. 

Distribution. — Specimens have been bred from larvae collected- at the following 
localities: Victoria: Franklin Riv. (6.3.53), Little Riv. and W. Tree Creek (north 
from Buchan, 17.1.56), Tubbut (17.1.56), Nowa Nowa (4.12.57), all in Gippsland, 
Warburton (18.3.58), Lome (1955-1958), Meredith (11.3.54), all collected by the author; 
Grampians (26.9.53) (A. Neboiss). S. Queensland: Upper Cedar Cr. (19.11.57, B. N. 
Marks), Lamington National Park (16.2.58, I. C. Yeo), Canungra (13.3.53), Nambour 
(25.4.45, J. L. Wassell), Somerset Dam (29.10.46, J. L. Wassell). 

AiSDES TUBBUTIENSIS, 11. Sp. 

Types. — The type series was bred from larvae collected at Tubbut (E. Gippsland). 
All specimens have their associated larval and pupal skins. The holotype female, 
allotype male, six paratype males and six paratype females are in the collections of 



140 



NOTES ON AUSTRALIAN MOSQUITOES. IV, 




Fig. 4. — Acdes tubbutiensis, n. sp. Aclult : a, male terminalia ; b. basal lobe of coxite : 
c, harpago ; Pupa : d, dorsal view of terminal abdominal segmients ; e, part of cephalothorax. 
Larva : f, prothoracic setae ; g, head, terminal segments and mentum. 



BY N. V. DOBROTWORSKY. 141 

the National Museum, Melbourne. One paratype male and one paratype female are in 
each of the following collections: C.S.I.R.O., Division of Entomology, Canberra; School 
of Public Health and Tropical Medicine, Sydney; University of Queensland, Brisbane; 
British Museum (Natural History), London; U.S. National Museum, Washington. 

Distinctive Characters. — Adult: Narrow curved scales on vertex pale. Proboscis 
mottled with white scales on basal two-thirds. Integument almost black. Scutum 
clothed with narrow dark bronze scales. Scales in front of scutellum narrow. Femora 
and tibiae mottled. Sternites white scaled with more or less conspicuous median black 
spot, or some mottling and apical lateral spots. Ninth tergite of male with prominent 
lobes and a few stout setae. Larva: Head dark brown. Head setae 4, 5 and 6 arranged 
to form apices of a triangle; setae 4 lie between setae 5 and slightly in front of line 
drawn through their bases. Prothoracic setae 1 and 4, single. 

Holotype Female. — Head: Vertex with pale, narrow, curved scales. Upright forked 
scales pale in centre. Torus black with patch of white scales. First flagellar segment 
of antenna pale at base with patch of white scales; other segments black. Palpi black 
scaled with patch of white scales at base and at apex of segment 3; segments 2 and 3 
with a few pale scales above. Proboscis black with mottling of white scales on basal 
two-thirds. Thorax: Integument almost black. Scutum clothed with narrow bronze 
and black scales, becoming pale around bare area near wing roots, on margins of 
scutum; two lateral pale patches near mid-length; scales in front of scutellum narrow. 
Scutellum pale scaled. Anterior pronotum with pale and black bristles, narrow curved, 
and elongate pale scales. Posterior pronotum with patch of broad white scales below, 
broad black scales in middle and narrow curved pale scales above; a few pale broad 
scales scattered in upper part of black area. Pleura with usual patches of broad white 
scales and pale bristles. Wing length: 5-6 mm. Upper fork-cell less than twice length 
of stem. Stem of halteres pale, knob pale scaled, except top and underneath, which are 
black scaled. Leg: Femora black scaled with white mottling; basal half pale scaled 
posteriorly. Knee spots yellowish. Tibiae black, mottled and with few white scales on 
base. Tarsal segments 1-2 of fore legs, 1-3 of mid legs and 1-4 of hind legs with basal 
white bands; bands on segment 4, half length of segment. Abdomen: Tergites black 
scaled. Tergite 1 with few pale and black scales in middle, 2-4 with incomplete basal 
white bands and lateral spots, 5-7 with bands joining lateral spots. Tergites 6-8 with 
scattered pale scales, increasing towards segment 8. Sternites white scaled with black 
m.ottling in middle and small apical, lateral spots. 

Paratype Females. — The chief variations shown in a series of 11 females are: Size 
of pale spot on vertex may be reduced. Mottling of proboscis may be reduced to a few 
pale scales only. Mottling of hind tibia may be reduced. Basal bands on tergites may 
be reduced to small spots. Median black spot on sternites may be reduced; there may 
be only a few scattered black scales. 

Allotype Male. — This differs from the holotype as follows: Proboscis black with 
occasional pale scales. Palpi about as long as proboscis including labella; last two 
segments and apex of shaft with long dark hairs. Patches of white scales on base of 
segments 2-5. Torus black; flagellar segments pale with dark silky verticillate hairs. 
Fore tarsi with 2, mid tarsi with 3 and hind tarsi with 4 basal white bands. Wing 
length: 4-0 mm. Upper fork-cell one and one-third times as long as its stem. Abdomen: 
Tergite 2 with narrow basal band, 3-6 with wide bands joining lateral spots. Sternites 
black scaled, with elongate, white lateral spots and a few white scales in middle of 
apical border. Terminalia (Fig. 4, a, 6, c) of Ae. tubbutiensis is very similar to that of 
Ac. queenslandis and Ac. rupest^'is, but the styles are shorter — about one-third of the 
length of the coxite; first 4-5 setae of basal lobe about twice as long as those in 
queenslandis. 

Paratype Males. — The chief variations shown in a series of 11 males are: Proboscis 
may be entirely black. The white lateral spots on the sternites may be small; white 
scales may extend from the lateral spots to the apical border and along it. 



142 



NOTES ON AUSTRALIAN MOSQUITOES. IV, 



Larvae (Fig. 4, /, g).- — Head and siphon dark brown; body blackish. Head about 
three-fourths as long as broad. Antenna about half length of head, clothed with 
spicules; seta 1, 4-5-branched, arising at about half-way from base. Head seta 4, tiny, 
5-7-branched; 5, 5-7-branched; 6, 3-5-branched; 7, 7-10-branched; 8, single, rarely 
forked at apex; 9, 2-3-branched. Mentum with 10-12 lateral teeth on each side. 
Prothoracic chaetotaxy: seta 1 and 2 single; 3, 4-6-branched, as long as 2; 4, short, 2- 
branched; 5 and 6, single; 7, 3-branched. Eighth abdominal segment: seta 1, 4-5- 
branched; 3, 11-12-branched; 5, 6-7-branched; 2 and 4 single. Comb composed of 
large patch of fringed scales. Siphon index 3-1-3-7, mean 3-4; seta 1, 7-11-branched. 
Pecten of 20-30 spines, mean 25. Anal segment: saddle covering about half segment. 
Seta 1, single, may be 2-branched; 2, 6-12-branched; 3, single; 4 of 15-16 tufts. Anal 
papillae about as long as saddle. 

Biology. — Ae. tuhbutiensis breeds mainly in back water in creek beds and at times 
in rock pools more or less exposed to the sun. It avoids completely shaded pools (see 
Table 1). 

Table 1. 

The Habitats of Mosquito Larvae in Tubbut Area. 







AS. 


Ae. 


Ae. 


Ae. 


Cuhx 


Anopheles 


A. 


Date. 


Habitat. 


tubbuti- 
ensis. 


queens- 
landis. 


rupestris 


albo- 
annulatus. 


douglasi. 


stig- 
maticus. 


annulipes. 


17.1.56 


Little River, ground pool 


















exposed to the sun . . 


14 


— 


— 


— 


— 


— 


— 


20.1.58 


Tubbut, ground pool ex- 


















posed to the sun 


36 


1 


— 


— 


1 


— 


— 


15.3.58 


,, ,, ,, 


26 


— 


1 


— 


— 


— 


3 


16.4.58 


Snowy River, small grassy 


















ground pool . . 


2 


1 


1 


— 


— 


— 


2 


16.5.58 


Sandy Creek, shaded 


















ground pool 


2 


23 


— 


1 


— 


31 


— 


,, 


Sandy Creek, rock pool. 


















exposed to the sun, 50 


















yards further on 


10 


6 


38 


— 


— 


— 


— 


,, 


Dam Creek, ground pool, 


















short period of direct sun 


7 


2 


— 


— 


— 


5 


1 


,, 


Dam Creek, rock pool, ex- 


















posed to the sun 


11 


— 


7 








~ 



Habits. — Adults were not collected in nature and there is no information on their 
biting habits. It can be expected that it will bite man as readily as do other Victorian 
members of this complex. 

Distribution. — All known breeding places are situated in the north-eastern corner 
of Gippsland, in wooded country at an elevation of about 1,000 feet. This area has an 
average rainfall of 25-30 inches. 

Aedes sp. form A. 
This unnamed form, belonging to the alboannulatus complex, which has a white 
ring on the tibiae (Lee, Dyce and O'Gower, 1957), was found breeding alone in a 
rock pool with clean water and exposed to dispersed sunlight at Tubbut (17.2.56). It 
was breeding also at Little River (17.2.56) in a rock pool exposed to the sun, with 
slightly murky water and decayed leaves on the bottom; in this pool the dominant 
species was Ae. rupestris. 



The Status of Members of the Complex. 
Comparative studies of queenslandis and rupestris have shown that the two species 
differ in both their morphology and ecology. The ecological differences seem to be 
largely a matter of different temperature preferenda. 



BY N. V. DOBROTWORSKY. 



143 



During 1955-58, a study was made of larval breeding sites of the two species in 
the rocky valley of the Erskine River, near Lome. In this particular place the river 
widens, forming shallow rapids with numerous rocky pools. These pools provide 
permanent breeding sites for rupestris even during the summer when the water 
temperature may exceed 30°C.; it never makes use of the cooler pools in the surrounding 
forest. The forest pools, 100-200 yards away, are the permanent breeding places of 
queenslandis. This species also uses the exposed rock pools during the cooler months, 
but in summer its larvae are only occasionally found in them, and then in the deeper 
ones. 

The problem of why queenslandis larvae should be practically absent from the 
exposed pools during the summer has been investigated by means of laboratory 
experiments. 

Influence of loater temperature on oviposition. — Blood-fed females of queenslandis 
and rupestris were placed in cages of six cubic feet capacity containing two dishes 
(7" X 10-5") of water with wicks of filter paper covering the sides and providing 
oviposition sites. In one experiment, the temperature of the water in one dish was 
17-18°C., in the other 27°C.; in the second experiment the temperatures were 14-15°C. 
and 30°C. Air temperature in the cages was 17-22°C. (Table 2). 

Table 2. 
Number of Eggs Deposited in Water of Different Temperatures. 



Species. 




Temperature of Water. 




17-18° C. 


27° C. 


14-15° C. 


30° C. 


queenslandis 
rupestris 


150 


278 


563 




1347 



From Table 2 it is evident that queenslandis would not normally deposit eggs on 
the edges of exposed rock pools during the summer; rupestris, on the contrary, 
would prefer these sites. Certainly in nature, during cold spells or cool evenings, 
when the temperature of water falls, queenslandis may deposit some eggs on the edges 
of the pools; these eggs, however, may be exposed to direct sunlight later on, and if 
there is no rain for some days the fall in water level may result in the eggs being 
subjected to desiccation. 

Resistance of eggs of queenslandis and rupestris to desiccation. — Batches of 65 eggs 
of queenslandis and rupestris were placed on filter paper and exposed for different 
periods to a R.H. of 54-58% at a temperature of 20°C. In one experiment the eggs con- 
tained mature embryos, i.e., larvae ready to hatch; in a second experiment, immature 
embryos. After desiccation the filter paper bearing the eggs was immersed in water. 
The percentages hatching are shown in Table 3. 



Table 3. 
Ability of Eggs to Withstand. Desiccation. 





Exposure. 


Percentage Hatch. 




queenslandis. 


rupestris. 


Mature embryo 
Immature embryo 


2 days 
12 days 


1-5 
13-3 


95-4 
98-0 



It is clear that the fully embryonated eggs of queenslandis are unable to withstand 
even a brief period of desiccation. This, indeed, was apparent after 24 hours, for 
within that period 61-6% of eggs had collapsed and others had opened. Of the rupestris 
eggs, on the other hand, only 4-6% collapsed and none opened. The higher survival of 



144 



NOTES ON AUSTRALIAN MOSQUITOES. IV, 




queenslandis in the experiment using eggs witli young embryos can be attributed 
to the fact that these eggs did not open; the embryos were thus protected to some 
extent from desiccation. 

In another experiment 125 eggs of each species were placed on wicks of filter 
paper about half an inch above water level (17-18°C.). During the next three weeks 
56% of queenslandis eggs hatched and the larvae were able to wriggle over the moist 
filter paper into the water; riipestris eggs did not hatch. 

In nature, during the summer, the hatching of queenslandis eggs above water 
level could often be fatal for the emerging larvae, but with favourable meteorological 
conditions, some might succeed in making their way into receding rock pools. They 
would then be exposed to water temperatures of 30 °C. or more, which would probably 
be unfavourable for their development. 

Work on the thermal death points of larvae of these two species is in progress. 

As mentioned earlier, the alboannulatus group is distributed in the tropics and 
subtropics, and there can be little doubt that queenslandis and 7'upestris had their 
origin in the north. Both are sylvan species, but, as emphasized above, they have 
different temperature requirements. Ae. queenslandis became adapted to cooler con- 
ditions, and, having dispersed southwards, presumably along the ranges, is an abundant 
mosquito in cooler parts of south-eastern Australia. Ae. riipestris has been less 
successful in the south. Suitable breeding sites, that is, rock pools exposed to the 
sun, are not common in Victorian ranges, and, while large local populations of 
rupestris may be found, the distribution of this species is very patchy. 

It is clear then that the morphological differences between queenslandis and 
rupestris are accompanied by physiological and ecological differences. However, the 
decisive factor in determining taxonomic status is reproductive isolation. The 
queenslandis and rupestris are sympatric forms (the area of distribution of 7'upestris 
lies entirely within that of queenslandis) , but the examination of hundreds of specimens 
of both forms, from all known Victorian localities, has not revealed any intermediates. 
It is evident that the two forms are reproductively isolated and that rupestris should 
be regarded as a distinct species. 

Both species bite man readily, but rupestris ceases to do so earlier in the year 
than queenslandis; early in June, when the temperature was 17°C., there were no biting 
rupestris, but queenslandis was scarcely less active than in the summer. 

Ae. tuhbutiensis, as far aa we know, has a very restricted distribution; it has been 
recorded only in east Gippsland. Ae. tuhbutiensis probably is more closely related co 
queenslandis than to the other species. Unfortunately, not much is known about the 
biology and ecology of this form, but it is apparent that it is more adaptable to higher 
temperatures than is queenslandis. Ae. tuhbutiensis breeds mainly in ground pools, 
occasionally in rock pools, in the full sunlight or exposed to the sun for part of the 
day. Only once have adults of tuhbutiensis (two females) been bred out from larvae 
collected from a shaded pool. 

Examination of 121 adults and 70 correlated larval skins revealed only one female 
which could be regarded as an intermediate. Although tuhbutiensis and queenslandis 
both vary considerably, there is not a continuous gradation between them. 

If there were not reproductive isolation from queenslandis, intermediate forms 
should occur commonly. Mayr, Linsley and Usinger (1953) have stated, in relation to 
allopatric forms, "forms that hybridize only occasionally in the zone of contact are 
full species". The present author believes that this criterion can also be used for 
sympatric forms and thus can be applied in the case of tuhbutiensis. As tuhbutiensis 
is morphologically and ecologically distinct from queenslandis and reproductively 
isolated, it should be regarded as a separate species. 

The yellow form of queenslandis merges into typical queenslandis and only extreme 
variants can be distinguished with certainty. The progeny of the females of the yellow 



PROC. Linn. Soc. N.S.W., 1959. 



Plate i. 



*»*. 
.* 



V ^ 



■^ . 





, / 










*.* 



10 



11 




Breakdown of monad development in Lcucoi:ooon virf/atus. 



Proc. Linx. Soc. N.S.W., 1959. 



Plate js. 



w 



m 







# 



t 




,% 



^1^ 









•• 




A'ariable tetrad pollen in Astroloma (1-3), Acrntriclie (4), Lrncopixjnn (5) and BfachiiJoiiui (G). 



Proc. Ltxn. Soc. N.S.W., 1959. 



Platk 111. 




Pkoc. Lixx. Soc. N.S.W., 1959. 



Plate iv. 




BX N. V. DOBKOTWOESKT. 14& 

form are very variable; some are typical yellow form, others are indistinguishable 
from queenslandis. Apparently it is not reproductively isolated from queenslandis and 
as a sympatric form cannot be regarded as a distinct species. 

Acknowledgements. 
The author is grateful to Dr. F. H. Drummond for assistance in tjie preparation of 
the manuscript; he is also particularly indebted to Dr. E. N. Marks, University of 
Queensland, for helpful discussion and providing larval and adult material from 
Queensland, to Dr. P. F. Mattingly, British Museum, for comparison of specimens of 
Ae. queenslandis and Ae. rupestris from Victoria with the types of C. queenslandis 
Strickl. and G. demansis Strickl., and to Mr. Ernest Bass, Tubbut, for collecting larval 
material in the Tubbut area and to Mr. A. Neboiss, National Museum, Melbourne, for 
the gift of material. 

References. 
Edwards, F. W., 1924. — A synopsis of the adult mosquitoes of the Australasian region. Bull. 

ent. Res., 14 : 351-401. 
Knight, K. L., and Marks, E. N., 1925. — An annotated checklist of the mosquitoes of the 

subgenus Finlaya, genus Aedes. Proc. U.S. Nat. Mus., 101 : 513-574. 
Leb, D. J., DtcEj a. Li., and O'Gower, A. K., 1957. — Blood-sucking flies (Diptera) and myxo- 
matosis transmission in a mountain environment in New South Wales. Austr. J. Zool., 
5: 355-401. 
Matr, E., LiINsleTj E. G., and Usinger, R. L., 1953.^ — Methods and Principles of systematic 
Zoology, New York. 



PROCEEDINGS. LXXXIV, PART 1. 1959 



CONTENTS. 

Pages 

Presidential Address, delivered at the Eighty-fourth Annual General Meeting, 
25th March, 1959, by Dr. S. Smith-White: 

Summary of Year's Activities 1-4 

Pollen Development Patterns in the Epacridaceae, A Problem in 

Cytoplasm-Nucleus Interaction. (Plates i-ii; twelve Text-figures.) 8- 35 

Elections 4 

Balance Sheets for the Year ending 2Sth February, 1959 5-7 

Physiological Specialization of Melampsora lini (Pers.) Lev. in Australia. By 

H. B. Kerr. (Plates iii-iv; three Text-figures.) 36-63 

The Families of Cycads and the Zamiaceae of Australia. By L. A. S. Johnson. 

(Four Text-figures.) 64-117 

Root Parasitism in Atkinsonia ligustrina (A. Cunn. ex F. Muell.) F. Muell. 

By Barbara P. Menzies and H. S. McKee. (Fourteen Text-figures.) .. 118-127 

Descriptions of Two New Species of Curis and One New Species of Stiginodera 

(Buprestidae). By C. M. Deuquet. (Three Text-figures.) 128-130 

Notes on Australian Mosquitoes (Diptera, Culicidae). IV. Aedes alboannulatus 

Complex in Victoria. By N. V. Dobrotworsky. (Four Text-figures.) . . 131-145 



(Issued 28th October, 1959.) 



DC=s^ 



Vol. LXXXIV. ' No. 390. 

Part 2. 

THE 

PROCEEDINGS 

or THK 

LiMNEAN Society 



or 



New South Wales 



FOR THE TEAR 

1959 



Part 2 (pp. 146-294). 

CONTAINING PAPERS READ IN MAY-AUOUST 

With ten plates. 
[Plates v-xiv.] 




Stdsbt: 

pbiittd) ain> pubushsd fob the bociett bt 

AUSTRALASIAN MEDICAL PUBLISHING CO. LTD., 

Seamer and Arundel Streets, Glebe, Sydney, 

and 

SOLD BY THE SOCIETY, 

Science House, 157 Gloucester Street, Sydney. 



Registered at the Oeneral Post Office, Sydney, for transmission 
by post as a periodical. 



Agent in Europe: 
David Nutt, 41 Colebrooke Row, London, N.l. 



5 



DI=* 



I 



The Liimean Society of New South Wales 



LIST OF OFFICERS AND COUNCIL, 1959-60. 



President: 

T. G. Vallance, B.Sc, Ph.D. 

Tiee-Pr esidents : 

S. J. Copland, M.Sc. F. V. Mercer, B.Sc, Ph.D. 

Lilian Fraser, D.Sc. S. Smith-White, D.Sc.Agr. 

Hon. Treasnrer: 

A. B. Walkom, D.Sc. 

Hon. Secretaries: 

W. R. Browne, D.Sc; A. B. Walkom, D.Sc. 

Council: 

R. H. Anderson, B.ScAgr. S. Smith-White, D.Sc.Agr. 

W. R. Browne, D.Sc E. Le G. Troughton, C.M.Z.S., F.R.Z.S. 

A. N. Colefax, B.Sc. T. G. Vallance, B.Sc, Ph.D. 

S. J. Copland, M.Sc J. M. Vincent, D.Sc.Agr., Dip.Bact. 

Professor R. L. Crocker, D.Sc. A. B. Walkom, D.Sc. 

J. W. Evans, M.A., D.Sc, Sc.D. H. S. H. Wardlaw, D.Sc, F.R.A.C.L 

Lilian Fraser, D.Sc. Professor W. L. Waterhouse, C.M.G., 
F. V. Mercer, B.Sc, Ph.D. M.C., D.Sc.Agr., D.I.C. 

L V. Newman, M.Sc, Ph.D. A. R. Woodhill, D.Sc.Agr. 
Professor B. J. F. Ralph, B.Sc, Ph.D., 
A.A.C.I. 

Auditor: 

S. J. Rayment, F.C.A. (Aust.). 



NOTICE 

Sets of the Peoceedings of the Linnean Society of New South Wales (except 
First Series, Volumes I- VI, 1875-1881), and separate Parts, may be purchased from the 
Society, Science House, 157 Gloucester Street, Sydney. Separate Parts may be obtained 
also from David Nutt, 41 Colebrooke Row, London, N.l. The stock of First Series, 
Volumes I-VI, is strictly limited and some Parts of these Volumes are out of print. 

From 1st March, 1954, the price of each complete volume will be £3 10s., plus postage. 
(All previous price lists are cancelled.) 

Index to Volumes I-L of the Peoceedinqs [Issued 15th February, 1929]. Pp. 108. 
Price 5s. 

The Macleat Memoel&l Volume [Issued 13th October, 1893]. Royal 4to, ii to 308 
pages, with portrait, and forty-two plates. Price £1 Is. 

Descriptive Catalogue of Austealian Fishes. By William Macleay, F.L.S. [1881]. 
A few copies only. Two volumes and supplement Price £2 2s. 

The Teanbactionb of the Entomological Societt of New South Wales, 2 vols., 8vo. 
[Vol. I (complete in five parts, 1863-66), price 70s. net. Parts 2-5 10s. each; Vol. II 
(complete in five parts, 1869-73), price 30s. net, or single Parts 78. 6d. each.] 



SUBSCRIPTION: £3 10s. per anniim; postase 2s. 



BY N. V. DOBROTWORSKY. 145 

fonn are very variable; some are typical yellow form, others are indistinguishable 
from queenslandis. Apparently it is not reproductively isolated from queenslandis and 
as a sympatric form cannot be regarded as a distinct species. 

Acknowledgeynents. 
The author is grateful to Dr. P. H. Drummond for assistance in the preparation of 
the manuscript; he is also particularly indebted to Dr. E. N. Marks, University of 
Queensland, for helpful discussion and providing larval and adult material from 
Queensland, to Dr. P. F. Mattingly, British Museum, for comparison of specimens of 
Ac. queenslandis and Ae. rupestris from Victoria with the types of C. queenslandis 
Strickl. and C. demansis Strickl., and to Mr. Ernest Bass, Tubbut, for collecting larval 
material in the Tubbut area and to Mr. A. Neboiss, National Museum, Melbourne, for 
the gift of material. 

References. 
Edwards, F. W., 1924. — A synopsis of the adult mosquitoes of the Australasian region. Hull. 

ent. Res., 14 : 351-401. 
Knight, K. L., and Marks, E. N., 1925. — An annotated checklist of the mosquitoes of the 

subgenus Finlaya, genus Aedes. Proc. U.S. Nat. Mus., 101 : 513-574. 
Lee, D. J., DtcEj A. L., and O'Gower, A. K., 1957. — Blood-sucking flies (Diptera) and myxo- 
matosis transmission in a mountain environment in New South Wales. Austr. J. Zool., 
5 : 355-401. 
Mayr, E., Linslbt^ E. G., and Usinger, R. L.., 1953. — Methods and Principles of systematic 
Zoology, New York. 




14G 



OBSERVATIONS ON THE ECOLOGY OP THE PHASMATID CTENOMORPHODES 

TESSULATA (GRAY). 
By P. Hadlington* and F. HoscHKE.f 
(Plates v-vi; two Text-figures.) 
!, [Read 27th May, 1959.] 



Synopsis. 

The phasmatid, t'lenomorphodes tessulata (Gray, 1835) is a defoliator of trees in the 
coastal forests of northern New South Wales where it reached outbreak proportions during 
1955-56. Its known range is from Colombatti State Forest, near Kempsey, New South Wales, 
to Goodwood, near Maryborough, in Queensland. Outbreaks have only been recorded from 
forests in New South AVales. C. tessulata is not host specific since it has attacked a wide range 
of species of the genera Eucalyptus, Syncarpia, Acacia and Casuarina. 

Descriptions of ad'ults and eggs, and a key to the nymphal instars are given. The life 
cycle in the field, laboratory observations and experiments are described. Investigations show 
that cleptid egg parasites and disease of the phasmatid eggs are regulating factors. 

A similar stand composition occurs in all the areas of high phasmatid population. The 
effect of the defoliation on the management of forests is considered. 

It is suggested that the occurrence of forest fires is followed by an increase in the numbers 
of C. tessulata. The probable effect of fire on the cleptid parasites is also discussed. There 
appears to be a correlation between the large fires of 1951-52 and' the outbreak of C. tessulata 
in 1955-56. 



Introduction. 

Plagues of phasmatids have been recorded on the highlands of New South Wales 
over the past 75 years, and although it is likely that plagues occurred prior to 1880, 
this is the earliest date for which accurate records are available. The species concerned 
in these areas are Podacanthus wilkinsoni Macleay, 1881, and Didymuria violescens 
(Leach, 1814). More recently Ctenomorphodes tessulata (Gray) has been noted in large 
populations on the north coast of New South Wales. It is rather surprising that these 
have gone unrecorded, the first indication of their presence being extensive areas of 
defoliated forest trees. This can be partly attributed to the great diversity of form 
among the Phasmatidae and their resemblance to the vegetation upon which they feed. 

A recent paper by Key (1957) on kentromorphic phases in the Phasmatidae has 
enabled the phase patterns to be correlated with the population densities in the three 
species reaching plague proportions. 

The taxonomy and nomenclature of this and other species has been reviewed by 
Key (1957), and an application has been made to the International Commission on 
Zoological Nomenclature for the names to be added to the official lists of generic and 
specific names. 

The descriptions of C. tessulata by Gray (1835), Westwood (1859) and Bates (1865) 
were based on a small number of museum specimens available at that time. Bates 
leeords that collections available to him bore the data, Moreton Bay, Australia. A single 
specimen in the collection of the Department of Agriculture and Stock, Brisbane, 
Queensland, was collected at Moreton Island in 1916. Collections have been made at 
Byron Bay and Richmond Range State Forest in New South Wales and a small infesta- 
tion was recorded from Washpool State Forest in January 1955. 

* Entomologist, Forestry Commission of N.S.W. 
t Forester, Forestrj' Commission of N.S.W. 

Peoceeuings of the Linnean Society of New South Wales, 1959, Vol. Ixxxiv, Part 2. 






BY P. HADLINGTON AND F. HOsCHIvE. 147 

Prior to 1956, C. tessulata had not been recorded as a pest, but has since caused 
defoliation of forest trees in three widely separated areas of north-eastern New South 
Wales. A feature cf the outbreaks was that severe defoliation occurred during 1955-56 
in all three areas. Nymphs and adults which were collected in these areas exhibited the 
pattern of the high-density phase. Single specimens of nymphs from Wedding Bells 
State Forest near Woolgoolga in November 1952, and adults of the same generation 
collected in February 1953 were described by Key as having a low density phase pattern. 
No further material from this forest was studied until January 1956, when defoliation 
occurred over approximately 100 acres and individual small trees of Casuarina torulosa 
carried as many as thirty-six phasmatids. 

High populations were recorded from Toonumbar State Forest, near Kyogle, in 
January 1956, when a comparatively small area of forest was defoliated. Observations 
were not continued in this area as a Are occurred in December 1956. Widespread, but 
scattered, defoliation was observed on Tauban, Ingalba and Colombatti State Forests, 
near Kempsey, in February 1956. Defoliation in these State Forests was negligible 
during the 1956-57 summer, and it appears that the population is now at a very low 
level. 

Distribution and Hosts. 

Label data on specimens from several sources* have been used as a basis for 
establishing the known distribution of C. tessulata. Specimens have been collected from 
Colombatti State Forest near Kempsey, N.S.W., to Goodwood, near Maryborough, 
Clueensland, but distribution is probably more extensive. 

The localities from which large numbers of C. tessulata have been observed vary 
in altitude from 200 feet to 2,000 feet, while single specimens have been taken mainly 
in coastal areas of altitudes less than 200 feet. 

Very Low Density Populations: N.S.W. : Byron Bay; Queensland: Beechmont, 
Beerburrum, Burleigh Heads, Goodwood, Landsborough, Nerang, Petrie and Moretoii 
Island. 

Low Density Populations: N.S.W. : Washpool and Richmond Range State Forests. 

High Density Populations: N.S.W.: Toonumbar, Wedding Bells, Tanban, Ingalba 
and Colombatti State Forests. 

Specimens at the Division of Entomology, C.S.I.R.O., Canberra, were examined by 
Dr. K. H. L. Key, who gave their respective phase patterns. This information has beea 
related to the observations made in various outbreak areas in New South Wales and 
where accurate field data are available there is a correlation between the phase pattern 
and the abundance of the species. 

Observations on host relationships have been made in Toonumbar, Wedding Bells 
and Tanban State Forests, the three main outbreak areas. The nymphal stages were 
often found on many tree species, but only the characteristic defoliation or act of feeding 
was taken as establishing a positive host record. Thus many possible hosts have been 
rejected on the basis of inadequate evidence of attack. 

Field studies establish that the following tree speciesf normally provide food 
material for C. tessulata: Syncarpia laurifolia Ten., Eucalyptus paniculata Sm., E. 
propinqua Deane and Maiden, E. gummifera (Gaertn.) Hochr., E. maculata Hook., E. 
resiniiera Sm., E. punctata DC, E. triantha Link, E. pilularis Sm., Casuarina torulosa 
Ait., Acacia floribunda (Vent.) Willd. 

Specimens were reared in the laboratory by Key (1957) on Acacia mollissima Wilid. 
and Eucalyptus dives Schauer. Adult phasmatids, collected from Tanban State Forest 
during January 1957, were held at Sydney and fed on locally available material of 
Angophora costata Domin, Eucalyptus radiata Sieb. and Syncarpia laurifolia. Some 

* Division of Britomology, C.S.I.R.O., Canberra ; Department of Agriculture, N.S.W. ; 
Department of Agriculture and Stock, Queensland ; Forestry Commission of N.S.W. 

t Identifications of trees in the field have been made by local field officers. 



■]48 



ECOLOGY OF THE PHASMATID CTEXOilORPHODES TESSULATA, 



first instar nymphs which were collected in the field were reared entirely on Syncarpia 
laurifolia in the laboratory at Sydney. 

These observations indicate that C. tessulata is not specific in its feeding habits as 
tree species of different genera, namely, Syncarpia, Acacia, Caauarina and Eucalyptus, 
have been attacked. It is probable that this phasmatid will feed on other species. 

Other phasmatids which have occurred in high populations are more specific m 
their tastes, but C. tessulata has nevertheless exhibited selective feeding within the 




O Isolated Occurpcnoi 
e Low Dansihy Populoh.oo 
9 Hi^h D«n3lt^ RapuUhioo 



Text-fig. 1. — Occurrence.s of C. tessnlata. 



genus Eucalyptxis. E. microcorys was less preferred, for during 1956 it was observed 
that, while crowns of E. microcorys showed no defoliation which could be positively 
attributed to C. tessulata. those of E. inacxil,ata> immediately adjacent were completely 
defoliated. 

Biology and Descriptions. 

During 1956 and 1957 intensive observations on C. tessulata were made in Tanban 
State Forest and in the laboratory at Sydney. Although Tanban State Forest is the 
southernmost of the three outbreak areas, and well separated from the others 
(Toonumbar and Wedding Bells State Forests), differences in the number of develop- 
mental stages of the insect between the three areas are considered unlikely to occur. 

The data given hereunder, particularly in respect of the seasonal cycle, apply to 
material collected from Tanban State Forest. Investigations on Toonumbar or Wedding 
Bells State Forests have been confined to the collection of adult material and samples 
of litter from the forest floor. 



BY P. HADLINGTON AND F. HOSCHKE. 



149 



(a) Seasonal Cycle: First instar nymphs were observed during late August and 
early September 1956, and most emergences had occurred before the end of September. 
Some fourth and fifth instar nymphs were collected during late September. These were 
always taken in areas having a north-easterly aspect, and it is presumed early hatching 
occurred in these situations. In general, only small variations in the proportion of the 
different instars were observed in the main population. The young nymphs ascended 
the trees near their hatching site and commenced to feed. Adults first appeared during 
December 1956, the males reaching maturity one or two weeks prior to the females. 
The adults of the high population recorded in January 1956 were not observed after 
February 1956, oviposition having occupied a period of 6-8 weeks. 

Adults of the 1957 infestation did not survive more than a few weeks in the field, 
but many of those which were collected on 16th January 1957, and held in the laboratory, 
survived until May 1957. 

(Z)) Egg Studies: Eggs were collected from forest litter during April 1956, 
immediately after the 1955-56 generation. Further samples were examined in December 
1956 and April 1957, when it was anticipated that the 1956 hatchings had occurred. 
Litter samples were collected in April 1957, to assess oviposition of the 1956-57 genera- 
tion. The results of these examinations are expressed in Table 1. 

Table 1 

Egg Dissections, Ctenomorphodes tessiilata, Tanban State Forest, 





Results expressed 


as percentages of total number of e 


ggs collected. 




Date of 
Collection. 


Embryonic 
Development. 


No Visible 
Development. 


Parasitism. 


Deterioration. 


Empty 
Shells . 


April, 1956 . . 
December, 1956 
April, 1957 . . 


7-9 
0-8 
00 


7-3 
2-5 
0-0 


17-7 
7-4 
5-8 


25-7 
34-5 
27-0 


41-4 

54-8 
67-2 



Apparently some of the eggs laid in January and February 1956 hatched in August 
and September 1956, thus differing from those of P. wilkinsoni and D. violescens which 
usually hatch during the year following oviposition. A low percentage of eggs (2-5%) 
showed no development by December 1956, but these eggs appear to have deteriorated, 
or were parasitized later in the season. Some eggs laid by the specimens collected 
from Tanban State Forest in January 1957, and held at normal atmospheric conditions 
in Sydney, showed advanced embryonic development when examined in June 1957. Eggs 
of the same origin which were held in soil at 78°F. and at room temperature hatched 
during June and July of the same year. The embryonic period, including at least one 
diapause, may occupy eight months or even longer in the natural habitat. The glistening 
ai)pearance which is characteristic of the freshly laid egg disappears during subsequent 
life of the egg when it is in soil, but no external signs such as colour changes have been 
noted which may indicate embryonic development. 

A number of eggs from the 1956-57 generation were placed in moist soil. One batch 
was held at 76°F. while another was kept at room temperature and therefore subjected 
to diurnal variations. The latter yielded healthy first instar nymphs mainly early ir 
the mornings (1.00 a.m.-6.00 a.m.) and only occasional hatchings took place at other 
times during the day. At 76°F. small and weakened first instar nymphs emerged only 
at irregular intervals. When transferred to plants and held at room temperature, they 
lived only a short time and were not observed to feed. 

Once initiated the embryo develops very rapidly until it entirely occupies the 
chorion and it has proved difficult to obtain early developmental stages of the embryo. 
Of the eggs collected in April 1956, 7-7% were in an advanced stage until hatching took 
place in August and September 1956 (Table 1). The diurnal seasonal increase in 
temperatures appears to be an important stimulus to hatching, and there is evidence 
to suggest that the litter must be damp or wet before first instar nymphs can emerge 
satisfactorily. 



150 



ECOLOGY OF THE PHASMATID CTENOMORPHODES TESSULATA, 



Parthenogenesio has been recorded in the Phasmatodea by Salmon (1955) and 
Other workers. Females of C. tessulata are able to produce eggs without fertilization 
and embryonic development and subsequent hatching occur in a proportion of these 
eggs. 

(c) Nymphs: It has been deteimiined by Key (1957) that there are six nymphal 
instars for the males and seven for the females. Results obtained with specimens 
collected at Tanban State Forest during 1956-57 agree with those obtained by Dr. 
K. H. L. Key working in Canberra on laboratory insects. The predominant instars 
collected at approximately weekly intervals are expressed in Table 2. 

No overt gregariousness by the nymphs has been shown either in the field or the 
laboratory. The insects are most of the time in a resting state, even in dense popula- 
tions. Specimens in the laboratory have been observed to feed more frequently during 
the day than the night, but temperature and other climatic factors probably determine 
times of feeding in the field. After some ecdyses the phasmatids feed on their exuviae, 
sometimes completely devouring them. 

Table 2. 
Tanban State Forest. Collection of Nymphs. 



Date of 


Predominant 


Date of 


Predominant 


Collection. 


Instars. 


Collection. 


Instars. 


3.10.56 


I and II 


7.12.56 


VI 


26.10.56 


III and IV 


14.12.56 


VI 


2.11.56 


III 


21.12.56 


VI 


8.11.56 


III and IV 


28.12.56 


VI, VII, A 


16.11.56 


IV 


4. 1.57 


VII, A 


23.11.56 


IV and V 


11. 1.57 


VII, A 


30.11.56 


VI 


18. 1.57 


A 



id) Adults: The nymphs reached maturity during the latter part of December 
1956, in Tanban State Forest. First instar nymphs which had been collected from this 
State Forest in September 1956, and held under high density conditions at Sydney, 
became adults approximately one week earlier. Fifty adult males and females could be 
readily collected in the field during January 1957, but the insects had almost disappeared 
from the same area by 6th February 1957. During the previous year adults were present 
until the first week in March. Female specimens collected on 21st January 1957 were 
kept alive in the laboratory until June 1957, while males collected on the same date had 
all died by the end of April 1957. Adults of this sipecies may live from 4 to 9 weeks 
in outbreak areas, but can be kept alive for at least five months in the Laboratory. 

Observations made on Ingalba and Tanban State Forests in February 1956 and 
January 1957 indicate that the ratio of adult males to females is approximately 50:50. 
but in some sites females predominate. The ratio at emergence is not necessarily the 
same as that of the adults, for males become adults earlier, and they may die prior to 
the females. 

Laboratory female specimens of C. tessulata laid from 300-900 eggs over a period 
of five months. The daily oviposition rate varied from 4 to 10 per female each day. 

Copulation occurs soon after the adult stage is reached and approximately two 
weeks elapse before the females commece to lay. The males and females remain 
together for long periods and move around while feeding. Copulation occurs at irregular 
intervals, oviposition taking place between these acts. Neither adults, nor nymphs, 
exhibit any overt .gregariousness. 

The destructive potential of C. tessulata was compared with that of P. ivilkinsoni and 
7>. inolescens by isolating two males in one cage and two females of the same species 
in another for all three species. Chewed off foliage and the faeces of these insects 
were collected, separated, oven-dried and wieghed daily. The figures obtained were 
halved, to give a daily destruction rate of foliage for one specimen of each species. A 



BY P. HADLINGTON AND F. HOSCHKE. 



151 



daily maximum and an average daily figure for fourteen days collection are given in 
Table 3. 

The defaecation rates have not been interpreted statistically as insufficient material 
was available to design an experiment, with sufficient replications. It is apparent that 
C tessulata is as destructive as either P. wilTcinsoni or D. violescens, both in the quantity 
of food eaten and foliage otherwise destroyed. Extensive areas of forest have not been 
observed defoliated to the same extent as occurs with P. wilkinsoni and D. violescens, 
and this is partly explained by the adults of C. tessulata being comparatively short-lived 
in the outbreak areas. Phasmatids of the highlands are in the adult stage for approxi- 
mately three months, while C. tessulata has not been observed six to eight weeks after 
maturity. 

Table 3. 

Comparison of Destructive Potential of Three Plague Phasmatids. 

Results shown as grammes of oven-dry weight of material per insect per day. 



Phasmatid. 


ITaeces. 


Foliage. 


Total. 


P. wilkinsoni (male) 








Maximum 


092 


0-092 


0-184 


Average 


054 


044 


0-098 


P. wilkinsoni (female) 








Maximum 


0-296 


0131 


0-427 


Average 


0-188 


066 


0-254 


D. violescens (male) 








Maximum 


0-140 


0-140 


0-280 


Average 


0-084 


0-023 


0-107 


D. violescens (female) 








Maximum 


0-211 


196 


0-407 


Average 


0-156 


0-062 


0-218 


C. tessulata (male) 








Maximum (1 day) 


0-012 


— 


0-012 


Average (1 day) . . 


0-012 


— 


0-012 


C. tessulata (female) 








Maximum 


0-278 


0-337 


0-615 


Average 


0-176 


0-111 


0-287 



It has been possible, in the case of the three plague phasmatids studied by Dr. 
K. H. L. Key, to correlate the high density phase pattern with most field outbreaks. 
This should enable entomologists to approximate the density of a population from 
which one or two specimens have been collected from the field. Large populations of 
immature phasmatids during the last ten years have not always been recorded until 
defoliation has become apparent in spite of the fact that single specimens have beer 
collected and sent to entomologists for examination. Now, in the light of present 
knowledge, when nymphs show the medium to high density phase pattern, the popula- 
tion may warrant further investigation. 

The following descriptions which are intended for use in the field will serve to 
distinguish C. tessulata from the other two species of injurious Australian phasmatids. 
They are based on fourteen males and twenty-three females in the collection of the 
Forestry Commission of .New South Wales. 

Male (Plate v, fig 1): Head bearing two large compound eyes and three ocelli. 
Antennae 26-segmented. Pronotum approximately 3 mm. long, surface uneven, narrower 
than head. Mesonotum approximately 13 mm. long, two rows of prominent dorsal 
tubercles numbering in all from 8 to 13, several slightly raised lateral tubercles, 
narrower than the pronotum. Metanotum approximately 7 mm. long, without tubercles, 
wider than mesonotum. Prothoracic legs bearing no spines, femora of mesothoracic 
leg? carry spines along their length with two large spines at distal end. Spines 
evident along femora and tibiae of metathoracic legs and distal end of each. Tegmina 
brown, approximately 10 mm. long. Hindwings approximately 35 to 40 mm. long extend 
over fifth and sometimes part of sixth abdominal segments. Ten abdominal segments. 



152 ECOLOGY OF THE PHASMATID CTENOMOKPHODES TESSULATA, 

Genital lobes appear as enlargement on ventral surface of segments VIII and IX. 
Tenth segment modified to form claspers which bear spines. Cerci about i inch long, 
leaf-like. 

Female (Plate v, fig 2) : Head bearing two large compound eyes and three ocelli. 
Antennae 25-28-segmented, shorter than those of male. Pronotum approximately 5 to 7 
mm. long, no tubercles present, slightly narrower than head. Mesonotum 20 to 23 mm. 
long with two rows of dorsal tubercles and several lateral tubercles. Tubercles more 
prominently raised than in male. Metanotum 7 to 8 mm. long, non-tuberculate. Pronotum, 
mesonotum and metanotum of equal width. Prothoracic legs bearing spines along, and 
at distal end of femora. Mesothoracic legs with spines along femora and tibiae, as well 
as at distal end of these segments. Spines also present along and at distal extremities 
of femora and tibiae of metathoracic legs. Tegmina brown, approximately 13-15 mm. 
long. Hindwings 25 to 30 mm. long, apices extending almost to posterior margin of third 
abdominal segment. Ten abdominal segments. Operculum which covers genital valves 
arises from segment VIII, its apex extending beyond segment X and located between 
the cerci. Cerci about i inch long, leaf-like. 

General colour of the males and females light to dark brown. 

The Egg: In the Phasmatodea the egg may be regarded as a useful taxonomic 
accessory. Egg approximately 3 mm. long and 2 mm. in diameter at broadest point. 
Roughly elliptical in shape, somewhat wider in region of micropyle, distinctly flattened 
at both ends. When freshly laid, glistening black in appearance. Pale grey area 
extending over and around micropylar area. Smooth edged grey pigment appearing to 
overlie the black. Egg smooth, although microscopic examination shows it to be uneven. 
Black raised ridge in the shape of a Y surrounds the micropylar orifice at non-opercular 
end of egg. 

Key to the Instars of C. tessulata. 
Sex determination : 

1. Operculum or minute furrow* on abdominal .sternum VIII present femalo. 

2. Operculum of minute furrow on abdominal sternum VIII absent ' male. 

Male. 

1. Lateral and dorsal mesothoracic tubercles not raised: wing- rudiments not evident; antenna 

with 9 segments which are clearly separated', length approx. 12 mm Instar 1. 

Lateral and dorsal tubercles raised 2. 

2. Dorsal tubercles not black : wing rudiments evident ; antenna with 9 segments, several 

partially divided ; length approx. 20 mm Instar 2. 

Dorsal tubercles black a. 

3. Apices of ■ tegmen rudiments not extending over the metanotum: wing rudiments not in 

contact along tlieir inner margin ; antenna with about 18 distinct or partially divided 

segments ; length approx. 30 mm Instar 3. 

Apices of tegmen rudiments extending over the metanotum 4. 

4. Wing rudiments not in contact along their inner margins: antenna with about 19 distinct 

or partially divided segments ; length approx. 40 mm Instae 4. 

Wing rudiments in contact along their inner margins 5. 

5. Wing- rudiments not extending- to the posterior margin of abdominal segment 1 : apices of 

tegmen rudiments do not reach the anterior margin of wing rudiments ; antenna with 

about 22 segments ; length approx. 5.5 mm Instar 5. 

Wing- rudiments extending- beyond the posterior margin of abdominal segment 1 ; apices 
of tegmen rudiments almost reaching- the anterior margin of wing rudiments ; antenna 
with about 24 segments ; length approx. 70 mm Instar 6. 

Female. 

T. Lateral and dorsal mesothoracic tubercles not raised: wing rudiments not evident; operculum 
present as a transverse furrow on sternum VIII ; antenna with 9 distinct segments : 

length approx. 12 mm Instar 1. 

Lateral and dorsal mesothoracic tubercles raised 2. 

2. Dorsal tubercles not black : operculum present as a pointed triangular projection on sternum 
VIII, Text-fig. 2 ; antenna with 9 segments, several of which are partially divided ; length 

approx. 20 mm Instar 2. 

Dorsal tubercles black 3. 

* Minute transverse furrow is present only in Instar 1, and live or recently killed material 
is necessary to see this structure. 



BY P. HADLINGTON AND F. HOSCHKE. 



153 



Operculum not covering- anterior genital valves on sternum VIII, Text-fig. 2 : wing rudiments 
minute; antenna with about 18 distinct or partially divided segments; length approx. 

30 mm INSTAR 3. 

Operculum covering or almost covering- anterior genital valves on sternum VIII, 
Text-fig. 2 4 

Tegmen rudiments extending not more than half their length over anterior margin of 
metanotum ; antenna with about 19 distinct or partially divided segments; length 

approx. 45 mm Instar 4. 

Tegmen rudiments extending more than half their length over anterior margin of 
metanotum 5. 

Apex of operculum extending- half-way along- sternum IX, Text-fig. 2 : and partly covering 
posterior genital valves ; wing- rudiments not in contact along- their inner margins ; 
tegmen rudiments separated along inner margin ; antenna with about 22 segments ; 

length approx. 55 mm Instab 5. 

Apex of operculum extending- more than half-way along sternum IX 6. 







Text-fig. 2. — Nymphal characteristics. From left : abdominal sterna of female instars. 

Inner margins of wing rudiments not touching : operculum reaching posterior margin of 
segment IX, or almost reaching apices of posterior genital valve ; wing rudiments 
extending almost to base of metacoxae. Tegmen rudiments extending at least half 
their length beyond anterior margin of metanotum: antenna with about 24 segments; 

length approx. 75 mm Instar 6. 

Inner margins of wing rudiments touch ; operculum reaching to or beyond midpoint of 
segment X and exceeding- apices of posterior genital valve ; wing rudiments strongly 
tessellated, extending- to base of metacoxae ; apices of tegmen rvidiments almost reaching 
anterior margin of wing rudiments. Antenna with about 26 segments ; length approx. 
110 mm Instar 7. 



Mortality Factors. 

Several factors have caused, or have been suspected of causing, mortalities in the 
various stages of C. tessulata. 

(a) Temperature and Moisture: The effects of weather in the field have not been 
studied, all observations being made on insects in the laboratory. Nymphs were at 
first difficult to rear, but when water was sprayed onto the leaves, the first instar 
phasmatids were observed to consume some of it and mortalities were reduced. When 
nymphs were held in jars with an adequate supply of food they were reared through 



154 ECOLOGY OF THE PHASMATII) CTENOMOKPHODES TESSULATA, 

successive instars. A more successful method consisted of placing the recently hatched 
nymphs on young plants in nursery tubes, which were placed in a small container of 
moist soil, within a wire gauze cage approximately 24 in. > 18 in. x 18 in. The foliage 
was sprayed with water at least twice a day and the soil was watered on alternate days. 
Mortalities were low and nymphs were observed to feed during the third day. Key 
(1957) successfully reared the phasmatids from first instar nymphs by placing them in 
small wire mesh cages which were located over a container of water holding the leaves 
and covered with an inverted, loosely fitting plastic jar. An atomizer was used to spray 
the foliage twice a day during the early instars, and once a day during later instars. 

Recently emerged nymphs apparently require moist conditions, which would be 
supplied as rain and dew in the field. Most emergences from eggs in the laboratory 
occurred during the early hours of the morning and their moisture requirements would 
probably be fulfilled by condensation. Moisture is apparently less critical during the 
later instars and in the adult stage, but prolonged dry conditions, particularly if 
accompanied by moderate to high temperatures, could cause high mortalities among 
the early instars. 

First instar nymphs obtained from eggs held at 76°F. died a few days after being 
placed on food at temperatures of 55°F. to 65°F., but nymphs which hatched at tempera- 
tures of 55°F. to 65°F. and were placed with those hatched at 76°F. were reared 
successfully. Eggs held exposed under laboratory conditions of atmospheric humidicy 
and temperature failed to hatch, while those kept in moist soil hatched normally. 
However, lack of moisture did not inhibit embryonic development, so that under dry 
conditions high mortalities would occur at hatching. Conditions such as these would 
not occur frequently under field conditions. 

(b) Food: This species has been reared on Syncarpia laurifolia. Acacia mollissima 
and Eucalyptus dives, and a large percentage of the nymphs were reared to adults when 
any one of these host plants was used exclusively. In the field, recently emerged 
nymphs which ascend a tree, the foliage of which is not acceptable to them, may 
desiccate before a suitable host species is found. 

(c) Previous History of the Eggs: Some eggs which were obtained from the 195.5-5G 
generation showed no hatching although exposed to conditions which apparently 
stimulated hatching of the eggs obtained from the 1956-57 generation. Such factors as 
food requirements of the mature female, parthenogenesis and others which have not 
been elucidated, may affect the viability of the eggs. 

(d) CannihalisDi: Key (1957) records cannibalism as a possible mortality factor, 
and he states that chewed legs of recently moulted nymphs have been found in the 
rearing cages. When rearing specimens at Sydney, no cannibalism was noticed, but 
v/hen ecdysis was completed, freshly emerged nymphs were observed to consume their 
exuviae and frequently the exuviae of others. This has been observed before ecdysis 
was completed; thus a nymph which has not shed the exuviae from its legs may possibly 
have them chewed off by another nymph. This could account, to some extent, for 
instances of limb regeneration frequently encountered in this species. Cannibalism is 
not considered to be a mortality factor of laboratory or field specimens, even though 
most mortalities of nymphs in the laboratory occurred at ecdysis. 

(e) Fire: A large forest fire could be catastrophic to phasmatid populations, and 
it may destroy many insects and reduce their available food supply. A limited fire 
occurred on Toonumbar State Forest in December 1956, and insects which had emerged 
during the August or September may have been killed. This has been confirmed by 
observations since the fire. 

(/) Disease: During mortality studies of Podacanthus ivilkinsoni, Casimir and 
Edwards (1955 — unpublished report, Forestry Commission N.S.W. ) isolated a species of 
entomogenous fungus from nymphs and adults, but they state that this does not present 
suflScient evidence of pathogenicity. No instances of death attributable to pathogens 
have been determined, nor has it been possible to relate micro-organism activity with 
mortalities of C. tessulata. Nymphs have died when held at high humidities in the 



BY P. HADLINGTON ANIJ F. HOSOHKE. 155 

laboratory and, some time after death, a fungal growth spread over the surface of their 
bodies. This fungus was considered to be a saprophytic type. 

Dissections of eggs collected from the north coast outbreak areas showed a high 
mortality percentage from diseases of unknown origin (Table 1). Deterioration due 
to fungi, bacteria and physiological causes about which little is known at present lias 
affected 25-7%, 34-5% and 27-0% of total eggs collected in April 1956, December 1956 and 
April 1957, respectively. From eggs of C. tessiilata which showed deterioration, two 
fungi, Verticillium sp. and Penicilliuvi sp.,* have been isolated. Both are commonly 
found in the soil and it is unlikely that they would be the primary cause of the 
deterioration. 

ig) Predators: Insect predators have not been recorded attacking any stages of 
C. tessulata, although such attacks probably occur. Nymphs of the highland plague 
species have been attacked by Harpohittacus sp.t (Bittacidae-Mecoptera) in the Jenolan 
area. 

Birds feed on phasmatids, but observations only include crows feeding on D. 
violescens and P. u-Ukinsoni in the highland areas. No doubt C. tessulata also falls 
prey to birds, although no instance of this has been observed. The resemblance of 
phasmatid eggs to seeds is rather striking, as can be seen from Plate vi. 

Eggs of C. tesnulata are shiny black and when freshly laid are easily discernible in 
the forest litter, but when the eggs lose their lustre, they become difficult to locate, as 
soil adheres to their surface. Quail have been active in Tanban State Forest during 
the latter part of the outbreak, and it is possible that there is some association between 
the activity of these birds and the reduction in the number of eggs. 

(h) Parasites: Egg parasites of the genus Myrmecomimesis (Cleptidae-Chrysidoidea) 
have been described by Riek (1955). The material examined by him was mainly 
obtained from the Jenolan area where plagues of Didymuria violescens and Podacanthvs 
ivilkinso7Ji were present. Eggs of D. violescens from Bago State Forest have yielded a 
total parasitism of up to 9% during 1954-55. During 1952, parasitism of P. wilkinsoni 
eggs averaged as much as 7% from several collecting sites on the Jenolan area. 

The males of the several species of Myrmecomimesis found on the coast are winged 
while the females are wingless. Parasitized phasmatid eggs can be distinguished by 
the presence of small pits on the surface of the chorion, made by the female wasp 
when ovipositing. A small hole passes from at least one of these pits, through the 
vitelline membrane, causing the latter to become dark brown at the puncture. One 
egg may show several of these punctures, but only one parasite develops in each egg. 
Whether the pits are the result of exploratory efforts by the adult parasite in deter- 
mining a suitable oviposition site or whether several parasites have actually oviposited 
in the egg has not been determined. 

Phasmatid eggs which were laid during January and February 1956 were found to 
contain parasites when examined in April 1956. The parasite larva completely occupied 
the vitelline membrane indicating that development to this stage was comparatively 
rapid. The eggs were held in the laboratory until December 1956, when adult cleptid 
wasps emerged. Diapause may occur since the larvae remain in an advanced larval 
stage until November or December, although the phasmatid eggs are parasitized during 
January to April of the same year. 

Attempts to induce parasitism by exposing C. tessulata eggs to M. ruhrifemur under 
confined conditions were unsuccessful and a detailed study on the life cycle of 
Myrmecomimesis spp has not been made. 

The specimens of Loboscelidea sp.t were obtained shortly after eggs and forest litter 
were received from Tanban State Forest in April 1956. No emergences occurred later 

* Specimens examined by D. W. Edwards, Pathologist, Forestry Commission, N.S.W. 

t Observations by L. Mors, Forester, Oberon, 1955. 

t This species is being- described by Mr. E. F, Riek, Division of Entomolosy, C.S.I.R.O., 
Canberra. 



156. 



ECOLOGY OF THE PHASMATIU CTENOMORPHODES TESSULATA, 



iu the year, as was the case with Myrinecomimesis spp. This indicates tliat certain 
differences exist between tlie life cycles of the two parasites. Since Myi'mecomiimesis sp. 
is more abundant than Loboscelidea sp., it is probable that the former would account 
for more parasitism. 

Parasitism of C. tessulata eggs from the coast is given in Table 4. In addition, 
fluctuations in parasite numbers have been recorded for Tanban State Forest and these 

Table 4. 

Egg Dissection of April, 1956, Collection. 

Expressed as percentages of total eggs collected. 





Embryonic 


No 


Parasitism. 


Deteriorated. 


Empty 


State Forest. 


Development. 


Development. 






Shells. 


Toonumbar 


19-4 


13- 2 


20-6 


21-4 


25-4 


Wedding Bells 


2-0 


1-9 


12-4 


21-2 


62 ■ 5 


Tanban 


7-9 


7-3 


17-7 


25-7 


41-4 


Ingalba 


15-9 


16-4 


20-0 


240 


23-7 


Colombattl 


5-6 


8-1 


7-5 


48-2 


30-6 



are expressed in Table 1. Parasitism of C. tessulata eggs is greater than that of either 
P. wilkinso7ii or D. violescens in the highlands, and this has been verified by observa- 
tions on the forest litter in Tanban State Forest in February 1956, when the parasites 
were active. At no time have the parasites been observed to be so abundant in the 
highland areas as to be apparent in the litter. The parasite which has occurred in all 
outbreak areas and in greatest abundance is Myrinecomimesis ruhrifemur (Riek). The 
known occurrences of the other species are given in Table 5, but more extensive collec- 
tions would probably have given a greater distribution of the species involved. 

T.^^BLE 5. 

Occurrence of Cleptid* Parasites in Outbreak Areas. 



Location. 


Species. 


Toonumbar State Forest 


Myriiicroiiiiiiicxis rubrifemur (Riek). 


Wedding Bells State Forest 


M. nili.il.nnir (Riek). 


Tanban State Forest 


M. rubrifiiiiur (Riek). 




M. nigripedicel (Riek). 




Loboscelidea sp. 


Ingalba State Forest 


M. rubrifemur (Riek). 




M. niqripeiiicel (Riek). . 




M. bispinosa var. (Riek). 



* Determinations by E. F. Riek, Division of Entomology, C.S.I.R.O., Canberra. 

Discussion. 

(a) Effect of Outbreaks on the Stand Composition and Forest Management. 

The tree stand composition in all areas attacked by C. tessulata shows considerable 
similarity. Attack appears to be restricted to some of the drier type coastal hardwood 
forests of northern New South Wales. 

A typical stand in which the attack occurs consists of Eucalyptus punctata (Grey 
Gum), E. triantlia (White Mahogany), E. paniculata (Ironbark), E. maculata (Spotted 
Gum), E. gummifera (Red Bloodwood), Syncarpia laurifolia (Turpentine) and Stringy- 
bark. Always associated with attacked stands is an understory of Casuarina toriilosa 
(Forest Oak), and E. microcorys (Tallow wood) is often associated with them. Other 
species which are occasionally found in attacked areas are E. resinifera (Red Mahogany), 
E. umbellata (Gaertn.) Domin (Red Gum), E. saligna (Blue Gum), Tristania conferta, 
R. Br. (Brush Box) and Angopliora spp. (Wild Apple). The understory, particularly 
in the Kempsey area, consists mainly of Leguminosae. On Tanban State Forest, in the 
Kempsey district, phasmatids have attacked a stand of E. pilularis (Blackbutt), this 



BY P. HADLINGTON AND F. HOSCHKE. 157 

being associated with Grey Gum, Spotted Gum, White Mahogany, Ironbark, Stringybark 
and Turpentine. E. piZztZaris constituted more than 50% of the stand, and this is the 
only instance recorded where C. tessulata has attacked this species. 

As C. tessulata lias avoided E. microcorys in preference to other species of eucalypts, 
it was thought that this selectivity may have been accounted for on the basis of oil 
content of the leaves. A chemical examination of the leaves of favoured hosts and 
those of E. microcorys showed that the major oil constituents were very similar, and 
from these analyses both types of leaves should be equally acceptable. The minor 
constituents of the oils were not determined, and it is possible that these may account 
for the observed selectivity in the feeding habits. 

Attack by C. tessulata in plague proportions has not yet been recorded as recurrent. 
The plagues during 1955-56 severely or completely defoliated areas of up to 100 acres. 
In 1956-57 the phasmatid population was less, and by the mid-summer of 1956-57 all 
of the trees attacked the previous season had recovered and recognition of the areas 
which had been severely affected the previous year was difficult. Severe attack on 
isolated areas without further heavy infestations the following season would be unlikely 
to have any serious or lasting effect on the trees, while repeated defoliations would 
either kill or retard the growth of the preferred species, and Tallowwood may be 
favoured as against the more susceptible species. The removal of cover would enable 
a greater degree of light to penetrate to the forest floor, thus releasing suppressed 
lignotubers. This again would be likely to favour the development of tallowwood 
stands rather than the more susceptible species. 

The attacked stands contain only a small proportion of Tallowwood and it would 
be some time before it regenerated sufficiently to restock an area, particularly as the 
outbreak areas are of low site quality. When considering the merits of restocking with 
Tallowwood, it should be remembered that C. tessulata attacks many other valuable 
species. 

(b) Influence of Forest Fires. 

In 1951-52, fires burnt over large tracts of forest on the north coast, including those 
areas where outbreaks of C. tessulata were later to occur. The same areas were some- 
times burnt more than once during the same season, particularly on Tanban State 
Forest. Although data have been available on the extent, intensity and date of occur- 
rence of the fires, it was realized that the actual boundaries of the affected areas would 
be only approximate. For this reason the fire records have been related to phasmatid 
outbreaks only in a general manner. 

Fires were recorded from Tanban, Ingalba and Colombatti State Forests and 
adjoining private property between 26th August 1951 and 4th January 1952, but in the 
six years since that time no extensive burns have occurred. Observations during the 
phasmatid plague of 1955-56 revealed that high populations were retricted to definite 
areas, whilst a low-density population existed throughout the rest of the regions. These 
high populations have occurred in most instances close to or within the limits of 
areas burnt in 1951-52. 

Records of firas on Wedding Bells State Forest show that areas heavily infested 
by phasmatids in 1955-56 were burnt in 1951-52 during the period from 30th September 
to 26th October 1951, and local observations confirm these records. The phasmatid 
infestation of 1955-56 in Toonunibar State Forest was found to be present on a site 
which had not been burnt for at least eight years. A widespread fire did occur on 
23rd September 1951, but this was some distance from the outbreak area. These fires 
varied in intensity from light ground fires to crown fires, and within a comparatively 
small area damage would have been caused to the crown of some trees whilst those in 
close proximity remained relatively untouched. The leaf litter is destroyed only by the 
hotter burns. 

Nymphs of C. tessulata emerge in late August and early September when thej 
ascend the trees, usually to the more tender terminal shoots. The egg parasites appear 



158 ECOLOGY OF THE PliASMATlU CTEKOMOKPHOUES TESSULATA, 

to emerge mainly during December and February, when the eggs of C. tessulata are 
available for parasitism. 

Fires which occur in the spring and early summer months may destroy the 
parasitized eggs of C. tessulata on the forest floor, and if the Are were not intense, it 
is likely that the developing nymphs present in the crowns of the trees would be 
unaffected. The eggs from this generation hatch the following spring; thus an increase 
in population could occur over the succeeding years. The figures of parasitism in 
Table 1 indicate that the cleptid parasites are important mortality factors and then- 
destruction would remove their regulating influence on the phasmatid population, with 
a consequent rise in phasmatid numbers. 

Records of fires on Ingalba, Tanban. Colombatti and Wedding Bells State Forests 
appear to be related to phasmatid plagues, although there are some areas in these State 
Forests which have been attacked by phasmatids without having been affected by fire. 

The nymphal and mature phasmatids of the 1956-57 season disappeared prematurely 
from the outbreak areas, and this cannot be attributed to the activity of cleptid egg 
parasites which had greatly increased in number during the absence of fires. This points 
to the fact that at least one other major regulating factor, at present undetermined, was 
operating during the 1956-57 season. Climatic records for January and February 1957 
do not suggest climate as the main factor, for adult phasmatids held in cages at Sydney 
lived many months. No dead phasmatids were collected from the forest litter, nor were 
any observed to exhibit moribund symptoms which may indicate disease. 

Females of C. tessulata are flightless, and migration could not account for the 
decrease in population. 

While fires may have had an infiuence by temporarily reducing the egg parasite 
population, it is also suggested that fires either in or near plague areas have had their 
influence on the bird population. This may have resulted in an increase in the popula- 
tion of C. tessulata, by the regulating factors being partly removed. The occurrence of 
a univoltine life cycle and the high egg laying potential of C. tessulata have probably 
enabled plagues to occur within four years of the fires. 

After a period of five years high populations of C. tessulata are no longer present 
and it appears that there has been a return to a more stable environment. It will not 
be easy to test the hypothesis of predatism by birds without high populations of C. 
tessulata. As the practice of burning the litter on the forest floor in the spring and 
summer period may prolong phasmatid plagues by destroying the egg parasites, the 
lelative merits of controlled burning should be considered in relation to plagues of 
G. tessulata. The effects of fires on the environment of a forest are not always fully 
realized. 

Acknotvled gevie^its . 

In the preparation of this paper the advice, particularly concerning taxonomy, of 
Dr. K. H. L. Key of the Division of Entomology, C.S.I.R.O., was frequently sought. 

The senior author is indebted to his colleagues Mr. K. G. Campbell and Mr. K. M. 
Moore for their assistance in the egg dissection work and for their helpful criticism 
of the manuscript. The drawings of the female genitalia were also done by Mr. Moore. 

The photos which form Plate v, figs. 1, 2 3 and Plate vi are by Mr. R. Moulton, 
Forestry Commission of New South Wales. 

Mr. R. H. Luke of the Forestry Commission of New South Wales assisted in the 
correlation of fire and outbreaks, while Mr. F. R. Humphreys, also of the Forestry 
Commission of New South Wales, made the chemical analyses of the leaves of several 
host trees. Mrs. M. Casimir, Forestry Commission of New South Wales, has given advice 
from time to time during the investigations. 

The assistance of, the Forestry Commission's field staff is gratefully acknowledged. 

Eejerences. 
BATEiS, H. W.,, 1865. — Trans. Linn. Soc. London. 25: 347. 

Craighead, F. C, 1950. — Insect enemies of eastern forests. U.S.D.A. Misc. Publ. No. 657, p. 97. 
ri).\TON, C. B., 1952. — "Walkingstick damage in hardwood stand's and its control with D.D.T. 
aerial sprays. Progr. Rep. For. Insect Lab. Milwaukee, Wise, April. 



160 



AN ANNOTATED CATALOGUE OF DESCRIBED AUSTRALIAN TABANINAE 

(DIPTERA, TABANIDAE). 
By I. M. Mackerras, Queensland Institute of Medical Research, Brisbane. 

(Nine Text-figures.) 
[Read 27th May, 1959.] 



Synojjsis. 

The described Australian species and subspecies of Tabaninae are arranged according to 
the modern classification of the subfamily, the numbers recognized as valid' being- : Tabanus, 
20 ; Paracanthocera, 1 ; Chalyhosoma, 2 ; Cydistomyia, 20 ; Dasyhasis, 56. Six remain unrecog- 
nizable, three are removed to other subfamilies, and nine are excluded as having been recorded 
wrongly from Australia. New synonymy : 

In Tabanus : queenslandii Ferg. nee Ric. = australicus Tayl. ; geraldi Tayl. - concolor Walk. ; 
aprepes Tayl. = toivnsviUi Ric. ; diminutus Walk. = praepositus Walk. ; 7iir/ritarsis Tayl. = dorso- 
bimaculatus Macq. 

In Cydistomyia: postica (Hardy nee Wied. ) = avida (Big.); sanguinaria (Big.) = nigro- 
picta (Macq.). 

In Dasybasis : darwinensis (Tayl.) = elavicallosa (Ric); quadrata (Tayl.) = constans 
(Walk.); pseudopalpalis Ferg. & Hill = nemotubei-culata (Ric); adelaidae (Ferg. & 
Hill) - rufifrons (Macq.) ; bassii (Ferg.) r: tasmanica (Ferg.) = neoeirrus (Ric.) ; gerald- 
tonensis (Tayl.) = regis-georgii (Macq.); regis-georgii (auct. nee Macq.) = spadix (Tayl.); 
pseudobasalis (Tayl.) = postica (Wied.); oraria (English) = macrophthahna (Schin.) ; 
acutipalpis (auct. nee Macq.) = exulans (Erich.) ; whitei (Hardy) = nepos (Walk.) = hebes 
(Walk.) ; gentilis iniminutus (Hardy) ~ gentilis (Erich.) ; anteeedens (Walk., 1854 nee 
1848) = edentula (Macq.); flindersi (Ferg.) = anteeedens (Walk., 1848) - gregaria (Erich.); 
brevidentata (Macq.) = fratercula (Macq.) = circumdata (Walk.); edentula (auct. nee 
Macq.) = abstersa (Walk.) = acutipalpis (Macq.). 



The subfamily Tabaninae is characterized by lacking functional ocelli and apical 
spurs on the hind tibiae, and by having the third antennal segment composed of a 
basal plate and four (occasionally three) terminal annuli, the style of the male 
hypopygium truncate, and the caudal ends of the spermathecal ducts of the female 
provided with mushroom-like expansions. It contains a large number of intergrading 
forms, which are exceedingly difficult to classify at both specific and supra-specific 
levels. I have recently been able to study the types of nearly all the Australian species, 
and the present paper is designed to clarify their synonymy, which had inevitably been 
obscure, as many of the crucial specimens had not been available to workers in 
Australia. 

It will probably be a considerable time before a full revision of the subfamily can 
be completed, and the classification adopted here is, to some extent, tentative. The 
short definitions of the supra-specific divisions apply to Australian species only. They 
are intended primarily to indicate natural congeries of species, which may be raised 
or lowered in taxonomic status as subsequent investigations may indicate. Similarly, 
the species are arianged within the groups in a natural rather than alphabetical 
sequence, so that related forms can be compared. There are about 18 undescribed species 
(6 Cydistomyia, 12 Dasybasis) in the collections studied, but they do not affect the 
present discussion. 

This work would not have been possible without helpful cooperation from a great 
many authorities. The type specimens in Australia were studied through the kindness 
of the Directors and Entomologists of the Queensland Museum, Brisbane, the Australian 
Museum, Sydney, the School of Public Health and Tropical Medicine, Sydney, the 

Proceedings of the Linnean Society of New South Wales, 1959, Vol. Ixxxiv, Part 2. 



BY P. HADLINGXON AND F. IIOSCHKE. 159 

Ghay^ G. R., 1835. — S'i/noiisis of the species of insects belonging to the family of Phasniidae, 

44 pp. (London). 
Key. K. H. L., 1957. — Kentromorpliic phases in three species of Phasinatodea. Aust. J. Zool., 

5 (3) : 247-284. 
O'Connor, B. A., 1949. — Some insect pests of Tonga. Agric. J. Fiji, 20 (2). 
RiEK, E. F., 1955. — Australian Cleptid (Hymenoptera Chrysidoidea) egg parasites of 

Cresmododea (Phasmodea). Aust. J. Zool., 3 (1): 118-130. 
Salmon, J. T., 1955. — Tuatara, 5 (3) : 77-81. 
V/ESTWOOD, J. O., 1859. — Cat. Phas. Brit. Mus., p. 115. 

EXPLANATION OF PLATES V-VI. 
Plate V. 

1. Gtenomorphndt's tessulata. Adult male. 

2. Ctenomorphodes tessulata. Adult female. 

3. Myrmecomimesis sp. A parasite of Phasmatid eggs. 

4. An outbreak tirea. Tanban State Forest, February, 1957. E. maculata with under- 
storey of C. torulosa. 

Plate vi. 

Phasmatid eggs. Top row, left to right: Ctenomorphodes tessulata (Gray), Ctenomorpha 

chronus (Gray), Podacanthus ivilkinsoni Macleay, Podacanthus viridiroseus (Gi'ay), 

Podacanthus typhon Gray. Bottom row, left to right: Didymuria violescens (Leach), 

Acrophylla titan (Macleay), Extatosoma tiaratiivi (Macleay), Tropidoderus childreni (Gray). 



BY I. jM. MACKEREAS. 161 

Macleay Museum, Sydney, the National Museum, Melbourne, and the South Australian 
Museum, Adelaide. Mr. H. Oldroyd, of the British Museum (Natural History), compared 
many specimens with types, including the types of Erichson's species, which were 
borrowed for the purpose by Mr. H. F. Mattingly of the same institution. Mr. Oldroyd 
also provided facilities for me to work in his Department, as did Professor E. Seguy, 
of the Museum National d'Histoire Naturelle, Paris. Professor Fritz Pens, of the 
Zoologisches Museum, Berlin, and Dr. Max Beier, of the Naturhistorisches Museum, 
Vienna, also compared specimens with types, and sent me types on loan when any 
difficulties of interpretation arose. To all of these I would express my most grateful 
thanks. 

Species of which the types have been examined personally are indicated by an 
asterisk (*) in the body of the paper. f 

Tribe TABANINI. 
Basicosta covered with short, strong setulae, similar to those on the adjacent, 
swollen part of the costa (Mackerras, 1956, Fig. 6, C). This character seems to hold 
absolutely in the Australasian and Pacific faunas. The fronts of the females narrow 
towards the antennae, and their palpi are usually swollen at the base and taper to a 
relatively sharp point. 

Genus Tabanus Linnaeus. 
Type species: Tahanus dovinus Linn., Europe. No Australian synonyms, 
This faunal element is of strictly Oriental origin, and none of the species can be 
separated from the widespread genus Tabanus, as it is at present recognized. The 
eyes are hairy (pattern not known) in umbripennis, bare in all the other species; 
conspicuously banded in pallipennis, particaecus and ceylonicus, bicoloured brown and 
green in australicus, concolorous green or brown in the remaining species. 

The pallipennis Group. 
Vein i?., with strong appendix; frons of female relatively wide (index not more 
than 3); callus usually divided into two parts. 

Tabakus pallipennis Macquart. 

TahaniLs pallipennis Macquart, 1846. Dipt, exot., Suppl. 1, p. 32. Type J, from 
New Holland, not found in London or Paris. 

*Ta'banus rufi^iotatus Bigot, 1892. Mem. Soc. zool. Fr., 5, p. 673. Type J, from 
Australia, in the British Museum (Natural History). 

*Tabanus elesteem Summers, 1912. Ann. Mag. nat. Hist., (8), 10, p. 224. Cotype 5$, 
from Darwin, N.T., in the British Museum (Natural History). 

*Tabanus lineatus Taylor, 1913. Rpt. Aust. Inst. Trap. Med. for 1911, p. 65. Type $, 
from Bowen, N.Q., in the School of Public Health and Tropical Medicine, Sydney. Not 
T. lineatus Fabricius, 1781 (= giganteus Degeer, 1776), Nearctic. 

The type of pallipenyiis is evidently lost. It could not be found by Ricardo (1914), 
by Dr. P. Grenier, who^ searched for it in Paris for me last year, or by myself, in July- 
August, 1958. Under the circumstances, it is unfortunate that Hardy (1948) upset 
Ferguson and Hill's (1920) tentative identification of pallipennis as another species, 
and displaced the name rufinotattis under which this one was well known. Tabanus 
designatus Ricardo, 1913, from New Guinea, is an additional synonym. 

Distribution. — Northern Territory; north and east coastal Queensland; north 
coastal New South Wales. Also New Guinea. 

Tabanus pakticaecus Hardy. 
Tabanus particaecus Hardy, 1948. Proc. R. Soc. Qd., 58, p. 178. No type designated. 
Hardy proposed particaecus as a "new name" for the species which he considered 
Ferguson and Hill (1920) had wrongly identified as T. pallipennis Macq. In order to 

t The types of iJasybasis gentiUs (Erichson) and D. grecjaria (Erichson) were received 
from Professor Peus, and Mr. Oldroyd's interpretation of them confirmed, after this paper had 
gone to press. 



162 ANNOTATED CATALOGUE OF DESCRIBED AUSTRALIAN TABANINAE, 

avoid further confusion, I now propose to validate it by designating as lectotype a J 
from the original bred series from Eidsvold, S.Q., referred to by Hardy, and now in 
the School of Public Health and Tropical Medicine, Sydney. 

Distribution. — Southern Queensland; New South Wales, west of the Divide; north- 
western Victoria. 

Tabanus umbripennis Ricardo. 

*Tabanus cinereus Walker, 1848. List Dipt. Brit. Mus., 1, p. 167. Type $, from 
Western Australia, in the British Museum (Natural History). Not T. cinereus 
Wiedemann, 1821, Neotropical. 

Tabanus umbripennis Ricardo, 1915. Ann. Mag. nat. Hist., (8), 16, p. 283. Nom. nov. 

This species is, superficially, remarkably like the predominantly Holarctic genus 
Hybomitra. 

Distribution. — South-western Western Australia. 

The innotabilis Group. 
Vein i?4 rarely with short appendix; frons of female relatively narrow (index more 
than 3 ) ; callus single, usually elongate. 

(A) A black species, with shining subcallus and creamy white tarsi. 

Tabanus ceylonicus Schiner. 

Tabanus ceylonicus Schiner, 1868. Reise Novara Dipt., p. 93. Type J. from Ceylon, 
in the Vienna Museum; a J from Bramston Beach, N.Q., was sent to Dr. Beier, who 
reported that it agreed fully with the type. 

*Tabanus kershaivi Ricardo, 1917. Ann. Mag. nat. Hist.. (8), 19, p. 221. Type J, 
from Claudie R., N.Q., in the National Museum, Melbourne. 

The synonymy was known to Ferguson, but apparently not published until Oldroyd's 
(1949) revision of the Papuan species. 

Distribution. — Widely in the Oriental region; north to the Philippine Is.; east to 
New Guinea, New Ireland and the Solomon Is.; and in north-eastern Queensland as 
far south as Innisfail. 

(B) Species with a continuous, median, pale vitta on most or all abdominal tergites. 

Tabanus queenslandii Ricardo. 

*Tabanus queenslandii Ricardo, 1914. Ann. Mag. nat. Hist., (8), 14, p. 393. Type $, 
from N.E. Queensland, in the British Museum (Natural History). 

*Tabanus ivalte^-i Taylor, 1919. Peoc. Linn. Soc. N.S.W., 44, p. 48. Type J. from 
Hay, N.S.W., in the School of Public Health and Tropical Medicine, Sydney. 

The name queenslandii has usually been applied to another species (australicus) 
which is common in Queensland; but the above synonymy was published by Taylor 
(1926), and subsequently confirmed by Oldroyd (in lit.) and by examination of the 
types. 

Distribution. — Rare in north coastal Queensland; most specimens from western 
New South Wales. 

Tabanus strangmani Ricardo. 

*Tabanus strangmannii Ricardo, 1914. Ann. Mag. nat. Hist., (8), 14, p. 393. Type 
JcJ, from Darwin, N.T., in the British Museum (Natural History). 

*Tabanus mastersi Taylor, 1917. Peoc. Linn. Soc. N.S.W., 41, p. 754. Type J, from 
Bowen, N.Q., t^ from Mackay, N.Q., in the School of Public Health and Tropical Medicine, 
Sydney. 

Miss Ricardo's name is now emended to strangmani, in accordance with the 
"Copenhagen Decisions". Taylor (1913) first misidentified the species as gregarius 
Erichson, but later named it and labelled types. 

Distribution. — Northern Territory; eastern Queensland, from Moa I. to Eidsvold; 
western .New South Wales (Moree). 



BY I. M. MACKERRAS. 163 

Tabanus breinli Ferguson & Hill. 

*Tatianus hreinli Ferguson and Hill, 1922. Pkoc. Linn. Soc. N.S.W., 47, p. 255. 
Type 5, from Palm I., N.Q., in the School of Public Health and Tropical Medicine, 
Sydney. 

Distribution. — North Queensland (Palm I. and adjacent coast). 

Tabanus austealicus Taylor. 

*Tabanus australicus Taylor, 1919. Proc. Linn. Soc. N.S.W., 44, p. 53. Type 5, from 
Brisbane, Q., in the Queensland Museum, Brisbane. 

This species was misidentified as queenslandii by Ferguson (1920), and it has 
commonly been so labelled in Australian collections. 

Distribution. — Northern Territory; Queensland, from Gulf of Carpentaria and 
Cairns to Brisbane. 

Tabanus concolor Walker. 

■^Tabanus concolor Walker, 1848. List Dipt. Brit. Mus., 1, p. 179. Type 5- from 
New Holland, in the British Museum (Natural History). 

Tabanus geraldi Taylor, 1920. Proc. R. Soc. Vic, 32, p. 166. The type 2, from 
Bathurst I., N.T., was stated to be in the collection of G. F. Hill, but it could not be 
found in Mr. Hill's lifetime, and must be presumed lost. I therefore select as neotype 
a 2 in the School of Public Health and Tropical Medicine, Sydney, from Groote 
Eylandt, Northern Territory, N. B. Tindale, identified by Taylor and agreeing with 
his description. 

T. concolor has not been recognized previously in Australian collections. The type 
is rather a small (13 mm.), faded specimen, but there is no doubt of its identity. 

Distribution. — Northern Territory (Bathurst I., Groote Eylandt). 

(C) Brown to greyish species, usually with an abdominal pattern of pale or dark 
triangles or spots. 

Tabanus cinerescens Macleay. 

Tabanus cinerescens W. S. Macleay, 1826. In: King's Narrative of a Survey of the 
Intertropical and Western Coasts of Australia, 2, p. 467. Mr. J. R. Henry searched the 
Macleay Collection in the University of Sydney for me, but could find no type specimen. 

*Tabanus tetralineatus Taylor, 1913. Rpt. Aust. Inst. Trop. Med. for 1911, p. 68. 
Type 2, from Darwin, N.T., in the School of Public Health and Tropical Medicine, 
Sydney. 

It is desirable to stabilize the name cinerescens, which is well established in the 
literature. I therefore select as neotype a 2. fi'om Katherine R., Northern Territory, 
Brown, in the School of Public Health and Tropical Medicine, Sydney. This specimen 
agrees with females in the British Museum, determined by E. E. Austen, who appears 
to have been the first modern worker (Austen, 1914) to recognize the species, and also 
with Ferguson's and Taylor's concept of it. 

Distribution. — North-western Australia; Northern Territory. 

Tabanus innotabilis Walker. 

*Tabanus innotabilis^. Walker, 1848. List. Dipt. Brit. Mus., 1, p. 177. Type 2. from 
New Holland, in the British Museum (Natural History). 

*Tabanus kurandae Taylor, 1919. Proc. Linn. Soc. N.S.W., 44, p. 58. Type 2. from 
Kuranda, N.Q., in the School of Public Health and Tropical Medicine, Sydney. 

Miss Ricardo (1915a) wrongly listed dorsobimaculatus Macq. as a synonym of 
innotabilis, and was followed by Ferguson and Hill (1920) and Surcouf (1921). Taylor 
(1917a) misidentified specimens as duplonotatus Ric, and Ferguson and Hill (1920) 
corrected the error. T. daruensis Oldroyd, 1949, from New Guinea, is an additional 
synonym. 

Distribution. — North-western Australia; Northern Territory; Queensland, from the 
Gulf of Carpentaria and the Torres Strait Is. down the east coast to Rockhampton. 
Also New Guinea, Solomon Is., Santa Cruz Is. 



164 ANNOTATED CATALOGUE OF DESCEIBEI) AUSTRALIAN TABANINAE, 

Tabanus davidsoni Taylor. 

*Tabanus davidsoni Taylor, 1919. Peoc. Linn. Soc. N.S.W., 44, p. 65. Type J, from 
Mt. Tamborine, S.Q., in the School of Public Health and Tropical Medicine, Sydney. 

This is the southern representative of iyinotahilis; I believe that it may be regarded 
as at least subspecifically distinct. 

Distribution. — South coastal Queensland; north coastal New South Wales. 

Tabanus pakvicallosus Ricardo. 

*Tabanus parvicallosus Ricardo, 1914. Ann. Mag. nat. Hist.. (8), 14, p. 394. Type 5. 
from Moreton I., S.Q., in the British Museum (Natural History). 

Distr-ibution. — South-eastern Queensland; north coastal and inland New South 
Wales. 

Tabanus townsvilli Ricardo. 

*Tabaniis townsvilli Ricardo, 1915. Ann. Mag. nat. Hist.. (8), 15. p. 281. Type 2(^, 
from Townsville, N.Q., in the British Museum (Natural History). The original mis- 
labelling noted by Ferguson and Hill (1922) has been corrected. 

*Tabanus aprepes Taylor, 1919. Proc. Linn. Soc. N.S.W., 44, p. 56. Type $, from 
Kuranda, N.Q., in the School of Public Health and Tropical Medicine, Sydney. 

*Tabanus batchclori Taylor, 1919. Pkoc. Linn. Soc. N.S.W., 44, p. 58. Type $, from 
Batchelor, N.T., in the School of Public Health and Tropical Medicine, Sydney. 

The synonymy of aprepes with toionsvilli was suggested by Mr. Oldroyd, and 
confirmed when the type was examined. 

Distribution. — Northern Territory; coastal and inland Queensland; coastal and 
inland New South Wales; south-western Western Australia (possibly a distinct race). 

Tabanus notatus Ricardo. 
*Tabanus notatus Ricardo, 1915. Ann. Mag. nat. Hist.. (8), 15, p. 283. Type % from 
Ching Do, N.Q., in the British Museum (Natural History). 

This species is closely related to T. lenticulatus Oldr. of New Guinea. 
Distril)ution. — North coastal Queensland, from Cape York to Mackay. 

Tabanus sequens Walker. 
*Tabanus sequens Walker, 1848. List Dipt. Brit. Miis., 1, p. 178. Type ?, from Port 
Essington, N.T., in the British Museum (Natural History). 

Distribution. — Northern Territory; east coastal Queensland, from Moa I. to Brisbane. 

(D) Pale, unadorned species, with grey or fawn scutum, and predominantly 
yellowish abdomen. 

Tabanus peaepositus Walker. 

*Tabanus praepositus Walker, 1848. List Dipt. Brit. Mus., 1, p. 158. Type 5- from 
Port Essington, N.T., in the British Museum (Natural History). 

^Tabanus diminutus Walker, 1848. List Dipt. Brit. Mus.. 1, p. 183. Type J, from 
Port Essington, N,T.. in the British Museum (Natural History). 

*Tabanus obsciirimaculatus Taylor, 1919. Prog. Linn. Soc. N.S.W., 44, p. 51. Type $, 
from Darwin, N.T., in the School of Public Health and Tropical Medicine, Sydney. 

T. di7ninutus had not been recognized previously, but the type is in fair condition, 
and is clearly only a small specimen of praepositus; the description also agrees. 

Distribution. — North-western Australia; Northern Territory; north Queensland 
(Moa I. and Lockhart R.). 

Tabanus dorsobimaculatus Macquart. 

*Tabanus dorsobimaculatus Macquart, 1850. Dipt. exot.. Suppl. 4, p. 29. Type J, 
from E. coast of New Holland, in the Paris Museum. 

*Tabanus nigritarsis Taylor, 1913. Bpt. Aust. Inst. Trap. Med. for 1911, p. 67. 
Type 2, from Ching Do, N.Q., in the School of Public Health and Tropical Medicine, 
Sydney. 



BY I. M. MACKERRAS. 16& 

Miss Ricai'do (1915a J listed clorsoMmaculatus as a synonym of innotahilis. but the 
type, whicli is in fair condition, is undoubtedly the same as nigritarsis, and the 
description also agrees better with nigritarsis than with innotaiilis. 

Distfibution. — Northern Territory; north Queensland, from Cape York to Towns- 
ville. 

Tabanus nigrimanus Walker. 

*Tabanus nigrimanus Walker, 1848. List Dipt. Brit. Mus., 1, p. 183. Type J. from 
Port Essington, N.T., in the British Museum (Natural History). 

*Tahanus baclius Summers, 1912. Ami. Mag. nat. Hist.. (8), 10, p. 225. Two cotype 
22, from Darwin, N.T., in the British Museum (Natural History). 

*Tabanus daphoenus Taylor, 1919. Proc. Linn. See. N.S.W., 44, p. 54. Type 2. from 
near Darwin, N.T., in the School of Public Health and Tropical Medicine, Sydney. 

Walker's type is in good condition, and the others agree with it. The species is 
poorly represented in Australian collections. 

Distribution. — Northern Territory; north Queensland (Lockhart R. ). 

Tabanus obscurilineatus Taylor. 

*Tabanus obscurilineatus Taylor, 1919. Proc. Linn. Soc. N.S.W., 44, p. 50. Type J, 
from Darwin, N.T., in the School of Public Health and Tropical Medicine, Sydney. 

A small, yellow species, notable for its extremely small callus, but it shows no 
other resemblance to the predominantly Holarctic genus Atylotus. 

Distribution. — North-western Australia; Northern Territory; north Queensland 
(Townsville). 

Tribe DIACHLORINI. 

Basicosta without strong setulae, contrasting with the costa (Mackerras, 1956, 
Fig. 6, B). Fronts of females usually diverging towards the antennae, occasionally 
parallel; their palpi usually slender, tapering gently to a rounded end. 

Genus Paracanthocera Enderlein. 

Type species: Acantliocera australis Ric, Australia. 

Antennae much longer than head, first segment three times as long as wide; slender, 
Chrysops-Uke species, with bulging, shining face, and conspicuous, brown pattern on 
wings. I had thought (These Proceedings, 82, p. 291) that this genus was a synonym 
of the earlier Lissimas End., but the type of the latter (*fenestratus End., Celebes) 
proved to be a much less specialized insect. 

_^ Paracanthocera australis (Ricardo). 

* Acantliocera australis Ricardo, 1915. Ann. Mag. nat. Hist., (8), 16, p. 16. Type J. 
from Kuranda, N.Q., stated to be in the Berlin Museum, but the specimen in the British 
Museum (Natural History) is labelled as type. 

Distribution. — North Queensland (Kuranda and Cairns district). A second species, 
*parallelus (Walk.), is known from Batchian. 

Genus Ciialybosoma Oldroyd. 

Type species: Tabanus metallicus Ric, New Guinea. 

Metallic blue-green, calliphorid-like flies, with subcallus, parafacials and face rather 
bare and shining. 

Chalybosoma cyanea (Wiedemann). 

Tabanus cyaneus Wiedemann, 1828. Ausser. zweifl. Ins.. 1, p. 152. Type 2 in the 
Berlin Museum. Dr Pens has reported that a 2 from south Queensland, sent to him 
for comparison, agreed perfectly with the type, which is labelled "Nov. Holl., Melly.". 

*Tabanus cyaneoviriclis Macquart, 1850. Dipt. exot.. Suppl. 4, p. 31. Type 2. from 
"Tasmanie" (erroneous), in the Paris Museum. 

Except for its metallic coloration, this species might be included with almost equal 
propriety in the genus Cydistomyia. 

Distribution. — Central and south coastal Queensland; eastern New South Wales. 



166 ANNOTATED CATALOGUE OF DESCRIBED AUSTRALIAN TABANINAE, 

Ch.\lybosoma casuarinae English, Mackerras & Dyce. 

*C}ialyhosoma casuarinae English, Mackerras and Dyce, 1957. Proc. Linn. Soc. 
N.S.W., 82, p. 292. Holotype $, allotype cS' niorphotype larva, and pupal skins of types, 
from Merricumbene, N.S.W., in the Division of Entomology, C.S.I.R.O., Canberra. 

A remarkable species, which breeds in rot-holes in Casuarina trees. 

Distribution. — North and south coastal Queensland; south coastal New South Wales. 

Genus Cydistomyia Taylor. 
Type species: Cydistomyia doddi Tayl. (= alhitliorax Ric), New Guinea. 
Smooth-bodied, non-metallic species; eyes bare in both sexes; vein i?,, without 
appendix; frons of female relatively narrow (index usually greater than 4). 

Cydistomyia torresi (Ferguson & Hill). 

*Tabanus torresi Ferguson and Hill, 1922. Proc. Linn. Soc. N.S.W., 47, p. 257. 
Type 2, from Moa I., Torres Strait, in the School of Public Health and Tropical 
Medicine, Sydney. 

This species belongs to the Papuan laetiis group. It and jMlmensis form part vf 
a recent intrusion from New Guinea. 

Distrihution. — Torres Strait Is. 

Cydistomyia palmensis (Ferguson & Hill). 

*Tahanus palmensis Ferguson and Hill, 1922. Proc Linn. Soc. N.S.W., 47, p. 256. 
Type 5, from Palm I., N.Q., in the School of Public Health and Tropical Medicine, 
Sydney. 

Distrihution. — North Queensland, from Cairns to Townsville. 

Cydistomyia avida (Bigot). 

*Atylotus avidus Bigot, 1892. Mem. Soc. zool Fr., 5, p. 673. Type ?, from Australia, 
in the British Museum (Natural History). 

*Tahanus fuscipes Taylor, 1913. Rpt. Aust. Inst. Trop. Med. for 1911, p. 62. 
Type $c^, from Magnetic I., N.Q., in the School of Public Health and Tropical Medicine, 
Sydney. Not Tahamis fuscipes Ricardo, 1908, Ethiopian. 

Tabanus taylori Austen, 1914. Ann. Mag. nat. Hist., (8), 13, p. 265. Nom. nov. 

Tabanus posticus Hardy, 1944, nee Wiedemann, 1828. 

Hardy followed a suggestion by Ricardo (1915a), but the type of posticus has 
proved to be a Dasybasis of the vetusta group. 

Distribution. — Coastal Queensland, from Palm I. (abundant) to Brisbane (rare). 

Cydistomyia nigropicta (Macquart). 

*Tabanus nigropictus Macquart, 1855. Dipt, exot., Suppl. 5, p. 24. Type J, from 
"Inde" (erroneous), in the British Museum (Natural History). 

*AtyIotus sanguinarius Bigot, 1892. Mem. Soc. zool. Fr., 5, p. 675. Type J, from 
Australia, in the British Museum ( Natural History ) . 

Miss Ricardo (1915a) pointed out that nothing like nigropicta was known from 
India, whereas it was close to, if not identical with, the Australian sanguinaria. 
Comparison of the types shows that they are conspeciflc. 

Distribution. — South-eastern Queensland: north-eastern New South Wales. 

Cydistomyia pseudoakdens (Taylor). 

*Tabanus pseudoardens Taylor, 1913. Rpt. Aust. Inst. Trop. Med. for 1911, p. 66. 
Type $, from Kuranda, N.Q., in the School of Public Health and Tropical Medicine, 
Sydney. 

A red-brown species, which is related to C. lorentzi (Ric.) in New Guinea and 
hyperythrea in south-eastern Australia. 

Distribution. — North Queensland, from Cairns to Mt. Spec. 



BY I. M. MACKEEEAS. 167 

Cydistomyia hyperytheea (Bigot). 

*Atylotus hyj)erytlii-eus Bigot, 1892. Mem. So'C. zool. Fr., 5, p. 674. Type $, from 
Australia, in the British Museum (Natural History). 

Distribution. — South Queensland (Mt. Tamborine); eastern New South Wales, from 
Dorrigo to Sydney. 

Cydistomyia ditplonotata (Ricardo). 

*Tabanus duplonotatus Ricardo, 1914. Ann. Mag. nat. Hist., (8), 14, p. 396. Type $. 
from S. Queensland, in the British Museum (Natural History). 

Taylor (1917a) misidentified innotabilis as clujilonotatus. and later (1917Z), 1918) 
misidentified duplonotatus as parvicallosus. 

Distribution. — South-eastern Queensland; eastern New South Wales (Sydney). 

Cydistomyia laticallosa (Ricardo). 
*Tabanus laticallosus Ricardo, 1914. Ann. Mag. nat. Hist., (8), 14, p. 395. Type £, 
from Moreton I., S.Q., in the British Museum (Natural History). 

* Tabanus rufoabdominalis Taylor, 1917. Proc. Linn. Soc. N.S.W., 42, p. 525. 
Type JJ', from Stradbroke I., S.Q., in the Queensland Museum, Brisbane. 

Distribution. — South coastal Queensland; north coastal New South Wales 
(Richmond R. ). 

Cydistomyia laticallosa var. heroni (Ferguson). 

*Tabanus heroni Ferguson, 1921. Rec. S. Aiist. Mus.. 1, p. 372. Type J, from 
Dorrigo, N.S.W., in the South Australian Museum, Adelaide. 

The status of this form is puzzling. It seems to be consistently larger and more 
broadly built than laticallosa, and to link laticallosa with the normally very different 
victoriensis. I have used the old, vague term "var." to indicate lack of precise 
knowledge. 

Distribution. — South Queensland (McPherson Range) ; northern New South Wales 
(Dorrigo plateau and adjacent coast). 

Cydistomyia doddi (Taylor). 

*Tabanus doddi Taylor, 1917. Peoc. Linn. Soc. N.S.W., 41, p. 758. Type $, from 
Kuranda, N.Q., in the School of Public Health and Tropical Medicine, Sydney. 

This species is not to be confused with the genotype, *Cydistomyia doddi Taylor, 
1919, which is restricted to New Guinea. Taylor (1913) originally recorded it as 
abstersus Walk., but described it and labelled a type, after Austin (1914) had pointed 
out the error. It is remarkable for having a long dorsal process on the third antennal 
segment, almost like Diclielacera. C. doddi, magnetica, alternata and iventivorthi are 
best treated, either as distinct species, or as forming a north to south array of sub- 
species. The former arrangement is more convenient here. 

Distribution. — North Queensland, from Kuranda to Townsville. 

Cydistomyia magnetica (Ferguson & Hill). 

* Tabanus alter-natiis var. magneticus Ferguson and Hill, 1922. Proc. Linn. Soc. 
N.S.W., 47, p. 258. Type $, from Magnetic I., N.Q., in the School of Public Health and 
Tropical Medicine, Sydney. 

Distribution. — Coastal Queensland, from Magnetic I. to Rockhampton. 

Cydistomyia alternata (Ferguson & Hill). 

*Tabanus limbatinevris Macquart, 1850. Dipt, exot., Suppl. 4, p. 29. Type J, from 
E. coast of New Holland, in the Paris Museum. Not T. limbatinevris Macquart, 1847 
(see under "Excluded Species"). Macquart had evidently considered changing the 
later name, because he labelled the type in Paris "T. fuscinevris Macq., n. sp.", but that 
name had already been used for a Neotropical species. 

Tabanus macquarti Ricardo, 1915. A7171. Mag. nat. Hist., (8), 15, p. 277. Nom. nov. 
Not T. macquarti Schiner, 1868, Neotropical. 



168 ANNOTATED CATALOGUE OF DESCRIBED AUSTRALIAN TABANINAE, 

Tatanus aUernatus Ferguson and Hill, 1922. Proc. Linn. Soc. N.S.W., 47, p. 258. 
Nom. nov. 

Distribution. — South-eastern Queensland; north-eastern New South Wales. 

Cydistomyia wentworthi (Ferguson & Hill). 

*Tahanus wentworthi Ferguson and Hill, 1922. Proc. Linn. Soc. N.S.W., 47, p. 259. 
Holotype J from Leura, allotype S from Blue Mts., N.S.W., in the Australian Museum, 
Sydney. 

The frons index is slightly less than 4 in some specimens, but usually between 4 
and 4-5. 

Distribution. — New South Wales (Blue Mts.). 

Cydistomyia victoeiensis (Ricardo). 

*Tabanus victoriensis Ricardo, 1915. Ami. Mag. nat. Hist., (8), 15, p. 275. Type J, 
from Dandenong Ranges, Victoria, in the British Museum (Natural History). 

This is the southern and highland representative of the docldi-toentworthi series. 
Usually it is quite distinct, but some specimens vary towards wentworthi, and the 
frons index is also sometimes less than 4. 

Distribution. — South Queensland (Mt. Tamborine and McPherson Range); high- 
land and coastal New South Wales; Victoria. 

The remaining six described species provisionally included in Cydistomyia are 
small (8-12 mm.), and of uncertain relationships. C. silviformis is fairly typical, but 
the musgravii complex may lead into the microdonta group of Dasybasis, and brevior 
■ and griseicolor, in different ways, into the clavicallosa group. 

Cydistomyia atmophoka (Taylor). 
*Tabanus atmo'phorus Taylor, 1919. Proc. Linn. Soc. N.S.W., 44, p. 59. Type J, 
from Kuranda, N.Q., in the School of Public Health and Tropical Medicine, Sydney. 
This may be no more than a northern subspecies of musgravii. 
Distribution. — North Queensland, from Lockhart R. to Mackay. 

Cydistomyia musgravii (Taylor). 

*Ta.banus musgravii Taylor, 1918. Rec. Aust. Mus., 12, p. 64. Type 2- from 
Underbank, N.S.W., in the Australian Museum, Sydney. 

This species is remarkable in having a narrow green band across the eye of the 
female. The frons index is 3 to 4. 

Distribution. — South-eastern Queensland, principally in the mountains; New South 
Wales, in high country from Dorrigo to the Blue Mts., and on the coast near Sydney. 

Cydistomyia rivularis (Ferguson & Hill), 

*Tabamis pygnuieus Ferguson and Henry, 1920. Peoc. Linn. Soc. N.S.W., 44, p. 842. 
Type $, from Kendall, N.S.W., in the Australian Museum, Sydney. Not T. i)ygmaeus 
Williston, 1887, Nearctic. 

Tabanus rivularis Ferguson and Hill, 1922. Proc. Linn. Soc. N.S.W., 47, p. 264. 
Nom. nov. 

This form is typically smaller than musgravii (about 10 mm., as against 11-12), and 
has a shorter callus and shorter, paler antennae; but there is considerable variation, 
and I doubt that it can be maintained as distinct. 

Distribution. — South-eastern Queensland; north coastal New South Wales and 
Barrington Tops. 

Cydistomyia silviformis (Taylor). 

*Tabanus silviformis Taylor, 1919. Peoc. Linn. Soc. N.S.W., 44, p. 62. Type $, from 
Stradbroke I., S.Q., in the Queensland Museum, Brisbane. 

A small (8-10 mm.), brown species, with frons index 4 to 4-5 and banded abdomen. 
Known only from three females of type series. 

Distribution. — South Queensland. 



BY I. M. MACKEERAS. 169 

Cydistomyia beevioe (Walker). 

*Tabanus hrevior Walker, 1848. List Dipt. Brit. Mus., 1, p. 188. Type J, from 
Port Essington, N.T., in the British Museum (Natural History). 

Tabanus marginatus var.B Walker, 1848. List Dipt. Brit. Mus., 1, p. 189. $, from 
Port Essington, in the British Museum (Natural History). Miss Ricardo (1915a) 
recorded the synonymy. I did not see this specimen; the true *inarginatus (Walk.) is 
a Mesomyia (Chrysopinae) . 

*Tat)anus aneUosus Summers, 1912. Ann. Mag. nat. Hist.. (8), 10, p. 226. Three 
cotype 22, from Darwin, N.T., in the British Museum (Natural History). 

*Tat)anus australis Taylor, 1917. Proc. Linn. Soc. N.S.W., 41, p. 757. Type 2, from 
Stapleton, N.T., in the School of Public Health and Tropical Medicine, Sydney. 

*Tat>anus crypserythrus Taylor, 1919. Proc. Linn. Soc. N.S.W., 44, p. 60. Type 2c?. 
from near Darwin, N.T., in the School of Public Health and Tropical Medicine, Sydney. 

The enlarged upper facets of the eyes of the male are bare, and the frons index of 
the female is 4 to 4-5; but cell Ri is wide, and vein R, more or less angulate, though 
without definite appendix. 

Distribution. — Northern Territory. 

Cydistomyia geiseicolor (Ferguson & Hill). 

*Tabanus griseicolor Ferguson and Hill, 1922. Proc. Linn. Soc. N.S.W., 47, p. 262. 
Type 2. from Hughenden, N.Q., in the School of Public Health and Tropical Medicine, 
Sydney. 

A typical, small, greyish Cydistomyia, except for the wide frons, with an index of 
2 to 2-5. Known only from four females. 

Distribution. — North-western Queensland (Hughenden, Richmond). 

Genus Dasybasis Macquart. 

Type species : Dasybasis aijpencliculata Macq., Australia. 

Dolichajilxa Enderlein, 1930. Type species: Tabanus gregarius Erichs., Tasmania. 

Non-metallic species, of varied habitus from smooth to relatively broad and hairy; 
eyes hairy, at least on the upper facets of the male; vein R^ with appendix (except in 
the microdonta group); frons of female relatively wide (index usually less than 3). 

This genus consists, partly of a clearly definable faunal element, and partly of a 
series of groups of uncertain affinities but seeming to find closer relationship here than 
elsewhere in the tribe. Some of these may require subgeneric recognition when the 
whole complex has been worked out. As the synonymy is complicated, it will be 
necessary to include brief definitions of the species to which the names are applied. 

The microdonta Group. 

Small to large (10-16 mm.), dark, parallel-sided species, distinguished by the 
combination of a smoothly curved vein Ri, without appendix, and (usually) hairy eyes 
in both sexes. Males with eyes not markedly swollen and upper facets not conspicuously 
enlarged. Females with frons wide to moderately narrow (index 2 to 4-5) ; callus 
usually elongate and nearly full width of frons at base; terminal abdominal segments 
dorsoventrally compressed. 

This small group of four species (two undescribed) is difficult to place. It could 
have evolved from the victoriensis and musgravii stocks of Cydistomyia; or it may 
represent the New Zealand subgenus Dasybasis (Protodasyommia) in Australia. It 
seems better, at the moment, to include it here. 

Dasybasis miceodonta (Macquart). 

*Tabanus microdonta Macquart, 1847. Dipt, exot., Suppl. 2, p. 17. Type 2- from 
Tasmania, in the British Museum (Natural History). 

Tabanus wynyardensis Hardy, 1916. Paj}. Proc. R. Soc. Tas. for 1916, p. 269. 
Location of the type (^, from Wynyard, Tasmania, not known, but I have seen specimens 
from the type series. 



170 ANNOTATED CATALOGUE OF UESCRIBED AUSTEAHAN TABANINAE, 

A large (15-16 mm.), distinctive species, with wide frons (index 2-5 to 3) and 
relatively narrow callus. 

Distribution. — Victoria ; Tasmania. 

Dasybasis rainbowi (Taylor). 

*Tabanus rainbowi Taylor, 1918. Rec. Aust. Mus., 12, p. 66. Type (^, from King 
George Sound, W.A., in the Australian Museum, Sydney. 

A small (11-12 mm.), undistinguished species, with relatively narrow frons (index 
3-5 to 4-5). One of the new species (from Western Australia) is related; the other 
(from Mt. Kosciusko) is like a hairy C. musgraini with a wide frons (index 2 to 2-5). 

Distribution. — South-western Western Australia. 

The clavicallosa Groii-p. 

Small to medium-sized (8-13, occasionally 15 mm.), smooth, yellowish to greyish 
species, with unpatterned scutum, and abdominal pattern usually rather vague. Eyes 
of known males (five species) markedly swollen, with conspicuously enlarged, densely 
hairy, upper facets. Females with eyes bare; frons medium to rather narrow (index 
2-5 to 4-5), parallel to slightly diverging; callus narrower than frons, often reduced, 
sometimes absent; terminal abdominal segments dorsoventrally compressed. 

This is a predominantly northern group; but the presence of an appendix (often a 
long one) on Ri, the hairy eyes of the male, the usually rather wide frons of the 
female (index 2-5 to 3 in ten of the species), and the evident relationship with the 
ociilata group, suggest origin from Dasybasis rather than Cydistomyia stock. It 
extends into New Guinea (three species) and Bougainville (one species). 

Dasybasis griseoannulata (Taylor). 

*Tabanus griseoanniilatus Taylor, 1917. Proc. Linn. Soc. N.S.W., 41, p. 756. 
Type 5, from Brock's Ck., N.T., in the School of Public Health and Tropical Medicine, 
Sydney. 

An obscure, greyish species, with unusually narrow frons (index 4-5); known only 
from a single female. 

Distribution. — Northern Territory. 

Dasybasis angusticallus (Ricardo). 

*Tabanus angusticallus Ricardo, 1917. Ann. Mag. nat. Hist., (8), 19, p. 218. Type $, 
from Melville I., N.T., in the South Australian Museum, Adelaide. 

A small (9-10 mm.), grey species, distinguished by the short, linear callus on the 
lower part of the relatively narrow frons (index 3 to 3-5). 

Distribution. — Northern Territory. 

Dasybasis tryphera (Taylor). 

*Tal)anus trypherus Taylor, 1917. Proc. Linn. Soc. N.S.W., 41, p. 755. Type $, from 
near Dai-win, N.T., in the School of Public Health and Tropical Medicine, Sydney. 

A very small (8 mm.), greyish species, with relatively wide frons (index 2-5 to 3) 
and large, pear-shaped callus. Known only from two females, it may be an extreme 
variant of clavicallosa. 

DistribtUion. — Northern Territory (Darwin). 

Dasybasis clavicallosa (Ricardo). 

*Ta'banus clavicallosus Ricardo, 1917 (Feb.). Ann. Mag. nat. Hist.. (8), 19, p. 219. 
Type 5, from Milsoii I., N.S.W., in the British Museum (Natural History). 

*Tabanus darivinen.sis Taylor, 1917 (Apr.). Proc. Linn. Soc. N.S.W., 41, p. 758. 
Type 5, from Darwin, N.T., in the School of Public Health and Tropical Medicine, 
Sydney. 

*Tabanus griseus Taylor, 1919. Proc. Linn. Soc. N.S.W., 44, p. 55. Type 2- from 
Brisbane, Q., in the Queensland Museum, Brisbane. 



BY I. M. MACKERRAS. 171 

A small (9-12 iinn.), grey species, with frons index 2-5 to 3, and pear-shaped callus; 
eyes of J with just perceptible hairs. Ferguson and Hill (1922) established the 
synonymy of griseus; the type of darwinensis almost completely lacks the normal paier 
banding on the abdominal tergites, but is otherwise identical. 

Distribution. — Northern Territory (Darwin); coastal Queensland (Townsvilie, 
Brisbane); coastal New South Wales (Newcastle to National Park south of Sydney). 
Though the species is not uncommon, its distribution seems to be patchy. 

Dasybasis clavicallosa banksiensis (Ferguson & Hill). 

*Tahanus clavicallosus var. banksiensis Ferguson and Hill, 1922. Proc. Linn. Soc. 
N.S.W., 47, p. 262. Type J, from Moa (Banks) I., Torres Strait, in the National Museum, 
Melbourne. 

This form may prove to be a distinct species, when sufficient specimens are available 
for study. 

Distribution. — Northern Territory (Melville I.): north Queensland (Moa I.). Also 
New Guinea. 

Dasybasis neogermanica (Ricardo). 

*Tabanus neogermanicus Ricardo, 1915. Ann. Mag. nat. Hist., (8), 15, p. 283. 
Type 5, from Darwin, N.T., in the British Museum (Natural History). 

*Tabanus fugitivus Taylor, 1919. Proc. Linix. Soc. N.S.W., 44, p. 61. Type $, from 
Darwin, N.T., in the School of Public Health and Tropical Medicine, Sydney. 

*Tabanus hilli Taylor, 1919. Proc. Linn. Soc. N.S.W., 44, p. 64. Type ?, from 
Darwin, N.T., in the School of Public Health and Tropical Medicine, Sydney. 

One of the larger species of the group (11-14 mm.); yellowish-brown, with frons 
index 2-5 to 3, long, narrow callus, and well-defined abdominal bands. 

Distribution. — North-western Australia; Northern Territory. 

Dasybasis germanica (Ricardo). 

*Tabamis germanicus Ricardo, 1915. Ann. Mag. nat. Hist., (8), 15, p. 282. Type $, 
from Cairns, N.Q., in the British Museum (Natural History). 

A small (7-11 mm.), variable, fawn and brown to yellowish-brown species, with 
wide third antennal segment, frons index 2-5 to 3-5, pear-shaped callus, and deep brown 
to blackish fore femora. 

Distribution. — Northern Territory; eastern Queensland, from Torres Strait Is. 
(Saibai, Moa) and Cape York to Gladstone. Also New Guinea. 

Dasybasis constans (Walker). 

*Tabanus constans Walker, 1848. List Dipt. Brit. Mus., 1, p. 186. Type J, from 
New Holland, in the British Museum (Natural History). 

*Tabanus quaclratus Taylor, 1919. Proc. Linn. Soc. N.S.W., 44, p. 52. Type $, from 
Darwin, N.T., in the School of Public Health and Tropical Medicine, Sydney. 

*Tabanus minor Taylor, 1919. Proc. Linn. Soc. N.S.W., 44, p. 64. Type 5, from 
Darwin, N.T., in the School of Public Health and Tropical Medicine, Sydney. Not 
T. minor Macquart, 1850, Neotropical. 

Tabanus minusculus Ferguson and Hill, 1920. Proc. Linn. Soc. N.S.W., 45, p. 466. 
Nom. nov. Not T. minuscMlus Hine, 1907, Nearctic. 

Tabanus palnierstoni Ferguson and Hill, 1922. Proc. Linn. Soc. N.S.W., 47, p. 263. 
Nom. nov. 

A small (7-11 mm.), greyish fawn, rather variable species, with frons index 2-5 to 3, 
callus usually oval and tapering to a point above, entirely yellowish antennae and 
femora, and banded abdomen. The synonymy of minor was established by Taylor 
(1926) on the authority of Austen; quadratus differs only in having the frons usually 
a little wider, and the callus oblong rather than oval. 

Distribution. — North-western Australia; Northern Territory. 



172 ANNOTATED CATALOGUE OF DESCRIBED AUSTRALIAN TABANINAE, 

Dasybasis PARVA (Taylor). 

*Tabanus parvus Taylor, 1913. Rpt. Aust. Inst. Trop. Med. for 1911, p. 69. Type J, 
from Darwin, N.T., in the School of Public Health and Tropical Medicine, Sydney. 

D. parva, neopalpalis and pseudocallosa are small (8-10 mm.), obscure, yellowish 
species, of which only a few specimens are known. The frons index of parva is 3-5, 
and the callus is light brown, wedge-shaped. 

Distribution. — Northern Territory. 

Dasybasis NEOPALPALIS ( Ferguson & Hill). 

*Tahanus palpalis Taylor, 1919. Proc. Linn. Soc. N.S.W., 44, p. 66. Type 2, from 
Batchelor, N.T., in the School of Public Health and Tropical Medicine, Sydney. Not 
T. 2Mlp(ilis Ricardo, 1911, Oriental. 

Tabanus neopalpalis Ferguson and Hill, 1920. Proc. Linn. Soc. N.S.W., 45, p. 465. 
Nom. nov. 

Frons index 2-5 to 3; callus dark brown, small, rounded. 

Distributiooi. — Northern Territory. 

Dasybasis pseudocallosa (Ferguson & Hill). 

*Tabanus ijseudocallosus Ferguson and Hill, 1922. Proc. Linn. Soc. N.S.W., 47, 
p. 254. Type J. from Darwin, N.T., in the National Museum, Melbourne. 

Frons index 2-5; callus light yellow, triangular, resting on the shining yellow 
subcallus; costal cell brown. 

Distribution. — Northern Territory; north Queensland. 

Dasybasis nemotuberculata (Ricardo). 

*Tabanus nemotuberculatus Ricardo, 1914. Ann. Mag. nat. Hist., (8), 14, p. 388. 
Type 2, from Cape York, N.Q., in the British Museum (Natural History). 

Tabanus pseudopalpalis Ferguson and Hill, 1922. Proc. Linn. Soc. N.S.W.', 47, p. 252. 
The type 2 was stated to be in Mr. Hill's collection, but it cannot now be found. I 
therefore designate as neotype' the remaining 2 of the type series, from Batchelor, N.T., 
G. F. Hill, No. 1405, now in the collection of the School of Public Health and Tropical 
Medicine, Sydney. This may be the original type, unlabelled; but one cannot be sure, 
because the field observation numbers included series taken at the same time and place. 

A small (10-11 mm.), slender, yellow-brown species, with rather narrow frons 
(index 3 to 3-5), only a variable indication of a light yellow callus, the costal ceU 
yellow or brown, and the radial cells sometimes darkened also. Northern Territory 
specimens (pseudopalpalis) have a wider antennal plate and less darkened wings than 
those from Queensland {nemotuberculata), and subspecific separation may ultimately 
be desirable. 

Distribution. — Northern Territory; north Queensland (both sides of Cape York 
Peninsula). 

Dasybasis spatiosa (Ricardo). 

^Tabanus spatiosus Ricardo, 1915. Ann. Mag. nat. Hist., (8), 15, p. 288. Type 2. 
from Stannary HilJs, N.Q., in the British Museum (Natural History). 

Distinguished from nemotuberculata by being somewhat longer (11-12 mm.) and 
paler, with a somewhat wider frons (index 2-5 to 3), and a more definite, brown callus. 

Distribution. — North Queensland. 

Dasybasis nemopunctata (Ricardo). 

*Tabanus nemopunctatus Ricardo, 1914. Ann. Mag. nat. Hist., (8), 14, p. 388. 
Type 2. from Dunk I., N.Q., in the British Museum (Natural iHstory). 

*Tabanus aurihirtus Ricardo, 1915. Ann. Mag. nat. Hist., (8), 15, p. 290. Type 2<S' 
from Townsville, N.Q., in the British Museum (Natural History). 

^Tabanus hackeri Taylor, 1917. Proc. Linn. Soc. N.S.W., 42, p. 522. Type 2- fi'om 
Bribie I., S.Q., in the Queensland Museum, Brisbane. 



BY I. M. MACKEREAS. 173 

A larger (10-15 mm.), stouter species than the preceding; pale grey to yellow, with 
rather vague or no abdominal pattern; frons index 3 to 4; no callus. Taylor (1919) 
misidentified a specimen from Stradbroke I. as spatiosa. Yellow specimens may be 
mistaken for vespiformis, but can be recognized by the flat terminal abdominal 
segments. 

Distribution. — East coast, from Cape York to northern New South Wales (apparently 
strictly littoral). 

Dasybasis ochreoflava (Ferguson & Henry). 

*Tabanus ochreoflavus Ferguson and Henry, 1920. Proc. Link. Soc. N.S.W., 44, 
p. 845. Type 2, from Kendall, N.S.W., in the Australian Museum, Sydney. 

A medium-sized (11-14 mm.), relatively robust species, distinguished from nemo- 
punctata by darker general colour, more definitely banded abdomen, and well-defined, 
naiTow, brown callus. 

Distribution. — South Queensland (Moreton I.); northern New South Wales 
(Kendall). 

The oculata Group. 

Differs from the clavicallosa group principally in having the eyes hairy in both 
sexes, the frons of the female usually wider (index 2 to 3), the general coloration more 
brownish, and scutum and abdomen with more definite pattern. The eyes of the known 
males (seven species) are usually not markedly swollen, nor with the upper facets 
conspicuously enlarged. Terminal abdominal segments of females dorsoventrally com- 
pressed, except regis-georgii, in which they are markedly narrowed. 

The distribution is essentially southern, only oculata and spadix extending into 
north Queensland. 

Dasybasis kufifrons (Macquart). 

*Tabanus rufifrons Macquart, 1855. Dijit. exot., Suppl. 5, p. 28. Type 5, from New 
Adelaide, N. Holland, in the British Museum (Natural History). 

*Tabanus meridionalis Ferguson, 1921. Rec. 8. Aust. Mus., 1, p. 376. Type $, from 
Adelaide, South Australia, in the South Australian Museum, Adelaide. Not T. 
7neridionalis Thunberg, 1827, from unknown country. 

Tabanus adelaidae Ferguson and Hill, 1922. Proc. Linn. Soc. N.S.W., 47, p. 264. 
Nom. nov. 

A pale, but patterned species, with an indefinite, light yellow callus. Macquart's 
type is somewhat greasy, which has made the frons seem reddish, but there is no 
doubt of its identity. Miss Ricardo (1917) misidentified this species as postponens. 
Distribution. — South Australia. 

Dasybasis neocireus (Ricardo). 

*Tabanus neocirrus Ricardo, 1917. Ann. Mag. nat. Hist., (8), 19, p. 223. Type $ 
(not so labelled), from Swansea, Tasmania, in the British Museum (Natural History). 

*Tabanus tasmanicus Ferguson, 1921. Proc. R. Soc. Vic, 33, p. 20. Type 5, from 
Dunally, Tasmania, in the Australian Museum, Sydney. 

*Tabanus bassii Ferguson, 1921. Proc. R. Soc. Tie, 33, p. 22. Type $, from Wilson's 
Promontory, Victoria, in the National Museum, Melbourne. 

A small-medium (10-13 mm.), variable, greyish species with medium frons (index 
about 3), small, club- or pear-shaped callus, and banded abdomen. There is no doubt 
that the type of neocirrus was mis-labelled, as pointed out by Ferguson (1921a) and 
Hardy (1934). The specimen in the British Museum agrees with Miss Ricardo's 
description, and is from her stated type locality; the one in the South Australian 
Museum, labelled as the type, is from South Australia, is evidently her second specimen, 
and is not conspecific with the first. The three types differ appreciably from one another, 
but I believe that they all fall within the range of variation of the species. 

Distribution. — Coastal New South Wales, south of Sydney; Victoria; Flinders I.; 
Tasmania. 



174 ANNOTATED CATALOGUE OF DESCRIBED AUSTRALIAN TABANINAE, 

Dasybasis neobasalis (Taylor). 

*Tabanus hasalis Walker, 1848. List Dipt. Brit. Mus.. 1, p. 182. Type 5. from New 
Holland, in the British Museum (Natural History). Not T. basalis Macquart, 1838, 
Oriental. 

Tabanus neobasalis Taylor, 1918. Rec. Aust. Mus., 12, p. 67. Nom. nov. 

A relatively robust (11-14 mm.), fawn-brown species, with distinctive abdominal 
pattern, medium frons (index 2-5 to 3), relatively wide, heart-shaped callus, and wide 
artennal plate. 

Distribution. — South Queensland (Eukey); mountain and coastal districts of New 
South Wales; Victoria. There is also a complex of related forms, some of which may 
require specific recognition, from Victoria, Tasmania and Western Australia. 

Dasybasis oculata (Ricardo). 

*Tabanus puslllus Macquart, 1855. Dipt, exot., Suppl. 5, p. 29. Type J, from Sydney, 
N. Holland, in the British Museum (Natural History). Not T. pusillus Macquart, 1838, 
China. 

Tabanus oculatus Ricardo, 1915. Ann. Mag. nat. Hist., (8), 16, p. 276. Nom. nov. 

*Tabanus kendallensis Taylor, 1919. Prog. Linn. Soc. N.S.W., 44, p. 68. Type 2. 
from Kendall, N.S W., in the School of Public Health and Tropical Medicine, Sydney. 

Distinguished from neobasalis by generally smaller size (10 mm.), somewhat wider 
frons (index 2-5), smaller, more rounded callus, and narrower antennal plate. I would 
have been doubtful of the value of these characters, but Miss English has informed me 
that the pupae also differ significantly. 

Distribution. — Eastern Queensland, from Cairns to the McPherson Range; coastal 
and mountain districts of New South Wales, from the Queensland border to Canberra. 

Dasybasis postponens (Walker). 

*Tabanus postponens Walker, 1848. List Dipt. Brit. Mus., 1, p. 179. Type J, from 
New Holland, in the British Museum (Natural History). 

A small-medium (10-12 mm.), nondescript, brown species, with banded abdomen; 
like a small acutipalpis, but immediately distinguishable by the somewhat wider frons 
(index 2-5 to 3), and narrowly oval callus with short extension. 

Distribution. — Coastal New South Wales, from Kendall to Nowra. There are what 
may be variants in South Australia and Western Australia, but their status is uncertain. 

Dasybasis Dixoni (Ferguson). 

*Tabanus dixoni Ferguson, 1921. Proc. R. Soc. Vic, 33, p. 23. Type 5, from Lake 
Hattah, Victoria, in the National Museum, Melbourne. 

A small-medium (10-12 mm.), fairly brightly patterned species, with a wider frous 
than postponens (index 2 to 2-5) and a markedly narrower callus than spadix. Specimens 
from South Australia were identified as regis-georgii (Ric. nee Macq. ) by Miss Ricardo 
(1917). 

Distribution. — Victoria; South Australia; south-western Western Australia. 

Dasybasis spadix (Taylor). 

*Tabanus spadix Taylor, 1917. Proc. Linn. Soc. N.S.W., 41, p. 761. Type ?, from 
Sydney, N.S.W., in the School of Public Health and Tropical Medicine, Sydney. 

*Ta,banus brisbanensis Taylor, 1917. Proc. Linn. Soc. N.S.W., 42, p. 526. Type ?, 
from Brisbane, S.Q., in the Queensland Museum, Brisbane. 

Tabanus regis-georgii Ricardo, 1915, and subsequent authors, nee Macquart. 

A small (9-11 mm.), compact, brightly patterned species, with wide frons (index 2) 
and large, heart-shaped callus. It was wrongly identified by Miss Ricardo (19156) as 
regis-georgii, and Australian workers have universally followed her determination, in 
spite of the fact that it is not known from Western Australia (see comment by 
Ferguson and Hill, 1920, p. 466). 

Distribution. — Eastern Queensland, from Cairns to Stradbroke I.; coastal New 
South Wales; Victoria (Mallacoota Inlet). 



BY I. M. MACKERRAS. 1'75 

Dasybasis spadix diemanensis (Ferguson). 

*Ta'banus diemanensis Ferguson, 1921. Proc. R. 8oc. "Fie, 33, p. 25. Type $ and 
allotype (S, from Bream Ck., Tasmania, in the Australian Museum, Sydney. 

This form has a slightly narrower frons and smaller callus than the mainland 
race, and the upper facets of the eyes of the male are only slightly enlarged over a 
relatively small area. It seems better to treat it as a subspecies than a full species. 
Taylor (1919) misidentified it as hrisbanensis. 

Distribution. — Victoria; Tasmania; one J from South Australia may belong here. 

Dasybasis regis-georgii (Macquart). 

*Taba7ius regis-georgii Macquart, 1838. Dipt, exot., 1 (i), p. 132. Type J. from 
King George Sound, W.A., in the Paris Museum. 

*Ta'banus geraldtonensis Taylor, 1919. Pkoc. Linn. Soc. N.S.W., 44, p. 70. Type $, 
from Geraldton, W.A., in the School of Public Health and Tropical Medicine, Sydney. 

A distinctive species, not only by the abdominal pattern, but by having the terminal 

"segments of the female compressed, and the eyes of the male distinctly swollen, with 

considerably enlarged upper facets, suggesting relationship with the macrophtliahna 

group. Macquart's specimen is old, rather wasted, and has only a label "Roi Georg.", 

but there is little doubt that it is the type; it is certainly the same species as Taylor's. 

Distribution. — South-western "Western Australia. 

The vetusta Group. 

Nearly concolorous, grey to yellowish, medium-sized to large (11-17 mm.) species. 
Eyes of males usually markedly swollen and with upper facets conspicuously enlarged, 
densely hairy. Females with eyes hairy in southern species, bare in northern ones; 
frons moderately wide (index 2-5 to 3), nearly parallel, with small callus or none; 
terininal segments of abdomen laterally compressed, producing a tent-like appearance 
of the cerci in end view (Text-flg. 1). 

This is a very distinctive group of six (two undescribed), coastal, sand-haunting 
species, which may merit higher taxonomic status, but its precise relationship to the 
macrophthalma group has still to be defined. 

Dasybasis caesia ("Walker). 

*Tabanus caesius "Walker, 1848. List Dipt. Brit. Mus., 1, p. 180. Type J, from 
unknown country, in the British Museum (Natural History). 

Tabanus leucopterus van der "Wulp, 1868. Tijdschr. Ent., 11, p. 98. Type 2, from 
Aru Is., in the Rijksmuseum Nat. Hist., Leyden. 

*Tabanus griseohirtus Taylor, 1917. Pkoc. Linn. Soc. N.S."W., 41, p. 753. Type J, 
from Melville I., N.T., in the School of Public Health and Tropical Medicine, Sydney. 

A grey species, with the eyes of the female bare, frons index 2-5, and no callus. 
The synonymy of leucopterus was established by Oldroyd (1949). 

Distribution. — North-western Australia, including Montebello I.; Northern Territory; 
Torres Strait Is. (Murray, Moa, Thursday); north Queensland, from Cape York to 
Townsville. Also New Guinea and Aru Is. 

Dasybasis vetusta ("Walker). 

*Tabanus vetustus Walker, 1848. List Dijit. Brit. Mus., 1, p. 179. Type $, from 
Swan R., "W.A., in the British Museum (Natural History)., 

An almost uniformly grey to yellowish-grey species, with eyes of the female hairy, 
frons index 2-5, and a small, dark callus. 

Distribution. — The typical race is restricted to south-western "Western Australia 
and South Australia. Specimens from "Victoria and Tasmania belong to a distinct 
subspecies with entirely dark third antennal segment. 

Dasybasis vespiformis (Ferguson & Henry). 
*Tabanus vespiformis Ferguson and Henry, 1920. Proc. Linn. Soc. N.S."W., 44, 
p. 840. Type $, from Kendall, N.S.W., in the Australian Museum, Sydney. 



176 



ANNOTATED CATALOGUE OF DESCRIBED AUSTKALIAN TABANINAE, 



A large (15-17 mm.), yellow species, with frons index 3, and only vague indications 
of a callus. The type specimen was damaged, and neither the name nor the original 
description gives a true impression of the species. 

Distribution. — Coastal New South Wales, from Kendall to National Park, south of 
Sydney. 

Dasybasis postica (Wiedemann). 

*Tahanus posticus Wiedemann, 1828. Ausser. zweifl. Ins., 1, p. 152. Type $, from 
Australasia, in the Berlin Museum. 

'^Tabanus pseudobasalis Taylor, 1918. Rec. Aust. Mus., 12, p. 68. Type 5, from 
King George Sound, W.A., in the Australian Museum, Sydney. 




(32>^ '^^'^C::^ aO^ 



Text-figs. 1-9. 
1 : End view of abdomen of Dasybasis vetusta (Walk.), $ from Western Australia. 2 : Same 
of D. liebes (Walk.), 5 from Tasmania. 3: Same of D. circumdata (Walk.), ? from Canberra. 
4 ; Eighth sternite and antenna of D. hebes, $ from S. Queensland. 5 : Same of D. circumdata, 
? from Dorrigo plateau. 6 : Antenna of Dasybasis exulans (Erich.), ? from Tasmania. 
7: Antenna otiD. acutipalins (Macq.), $ from Victoria agreeing with type of abstersa (Walk.). 
S: Same of typical ? from Tasmania. 9: Frons, antenna and palp of D. tasmaniensis (White), 
paratype 9- Scales: top, for eighth sternites ; lower left, for abdomens, antennae and palp; 
lower right, for frons. 



A medium-sized (11-14 mm.), plump, unadorned species, with olive scutum and 
yellow abdomen. The possibility that this species might be postica had not been 
considered until the type was received for study. 

Distribution. — South-western Western Australia; South Australia. 

Both the new species are robust and yellowish, like vespiforniis. One, with eyes 
bare in the female, is from north-western Australia; the other, with eyes hairy in 
both sexes, is from south coastal New South Wales and Victoria. 

• Tlie macrophthalma Group. 

Robust (13-16 mm.), rather flat-bodied, grey or brown species, with well-developed 
scutal and abdominal patterns. Byes densely hairy in both sexes; more or less swollen 
and with moderately to conspicuously enlarged upper facets in the males. Females 
with wide (index 2 to 3), almost parallel frons; callus usually full width of frons 
and restricted to its lower half; terminal abdominal segments narrowed, cerci tent-like, 
eighth sternite either small or pointed apically (Text-fig. 2). 

There are eight Australian species (three undescribed) in this group, and they have 
a rather characteristic appearance, which is contributed to by a deep subcallus and 



BY I. M. MAOKEERAS. 177 

smoothly bulging paiafacials and face in the females; in most species, too, the antennae 
are black, contrasting with the wide expanse of whitish tomentum below them. It is 
difficult to decide how far the characters they share with the vetusta group are 
expressions of relationship or of common adaptation to life in a sandy environment. 
On the one hand, they are reinforced by common, distinctive features in the known 
pupae (Miss English, personal communication); but, on the other hand, the adults of 
exvJans and its immediate allies seem to lead rather towards the appendiculata group. 
Three species are known outside Australia, one from New Guinea, one from the 
Admiralty to the Santa Cruz Is., and one from New Caledonia and Loyalty Isi. The 
eyes are bare in the females, but otherwise they are very like macrophthalma. 

Dasybasis macrophthalma (Schiner). 

*Tabanus macr-ophthalmus Schiner, 1868. Reise Novara Dipt., p. 83. Type cJ, from 
Sydney, N.S.W., in the Vienna Museum. 

*Tabanus orarius English, 1949. Peoc. Linn. See. N.S.W., 74, p. 154. Holotype J, 
allotype J, morphotype larval and pupal skins, from Narooma, N.S.W., in the Macleay 
Museum, University of Sydney. 

A distinctive, grey species, which breeds in the sand of ocean beaches. The type 
from Vienna is labelled "megalops" and "Sydney". It agrees perfectly with the descrip- 
tion of macrophtJiahna, down to individual details, such as the retracted abdomen, and 
there is no megalops in the Novara Diptera, so there is no doubt that it is the specimen 
described. It also agrees perfectly with recent males of oraria. 

Distribution. — South coastal New South Wales. 

Dasybasis albohiktipes (Ferguson). 

*Tabanus alboMrtipes Ferguson, 1921. Rec. 8. Aust. Mus., 1, p. 377. Type $, from 
South Australia, in the South Australian Museum, Adelaide. 

A brownish-grey species, which falls naturally between macrophthahna and cirrus. 

Distribution. — South Australia; south-western Western Australia. There is also a 
related, undescribed species in Western Australia. 

Dasybasis cirrus (Ricardo). 

*Tabanus cirrus Ricardo, 1917. Ann. Mag. nat. Hist.. (8), 19, p. ,222. Type J, from 
Milson I., N.S.W., in the British Museum (Natural History). 

*Tabanus robustus Taylor, 1919. Peoc. Linn. Soc. N.S.W., 44, p. 69. Type $, from 
Brisbane, Q., in the Queensland Museum, Brisbane. 

A large (16 mm.), dark greyish-brown species, with characteristic drop-shaped 
callus. 

Distribution. — East coastal, from Palm I. in north Queensland to the Hawkesbury R. 
in New South Wales. 

Dasybasis exulans (Erichson). 

*Tabanus exulans Erichson, 1842. Arch. Naturges., (1), 8, p. 270. Type J, from 
Tasmania, in the Berlin Museum. 

Tabanus circumdatus var. White, 1915, nee Walker. 

Tabanus acutipalpis auct. nee Macquart. 

This robust (15-16 mm.), dark, southern species was wrongly identified as 
acutipalpis by Ricardo (19156), Ferguson (1921a, 1221b) and Hardy (1934, 1939, 1948). 
The antennal plate is angulate dorsally in the type, but more usually rounded as in 
Text-figure 6. The eighth sternite is small, and more concave distally than in circum- 
data (Text-flg. 5). 

Distribution. — Eastern New South Wales (chiefiy montane); Victoria; Tasmania. 

Dasybasis hebes (Walker). 

*Tabanus hebes Walker, 1848. List Dipt. Brit. Mus., 1, p. 159. Type J, from 
unknown locality, in the British Museum (Natural History). 

*Tabanus nepos Walker, 1848. List Dipt. Brit. Mus., 1, p. 181. Type $, from 
unknown locality, in the British Museum (Natural History). 



178 ANIVOTATED CATALOGUE OF DESCRIBED AUSTRALIAN TABANINAE, 

Tabanus circumciatus White, 1915, nee Walker; also other authors in part (records 
from Tasmania and most, if not all, from South Queensland; others probably represent 
a mixture, with this species in the minority). 

Tabanus whitel Hardy, 1939. Proc. Linn. Soc. N.S.W., 64, p. 44. Proposed as a 
"new name" for Tabanus cireumdatus White, 1915, nee Walker, 1848, and no type 
indicated. In order to fix its identity, I therefore select as lectotype a J in the School 
of Public Health and Tropical Medicine, Sydney, from Mangalore, Tasmania, 18.1.1914, 
A. White, labelled "7'. cireumdatus Walk." in White's handwriting. 

A slightly smaller (13-15 mm.), duller species than the preceding; antennal plate 
almost as wide, but with sharp dorsal angle; most easily recognized by its distinctive 
eighth sternite (Text-figs. 2, 4). 

The types of hebes and nepos are large (15 mm.), brown females, with clearly 
visible, pointed eighth sternites. A female from Eidsvold, south Queensland, returned 
to Ferguson by Austen as agreeing with the type of cireumdatus. agrees very well with 
these two, but not so well with the type of cireumdatus. The type labels are very small, 
and it seems possible that Austen may have mistaken the specimen he picked out for 
comparison. 

Distribution. — South-eastern Queensland; eastern New South Wales; Victoria; 
Tasmania; South Australia; south-western Western Australia. 

The trilinealis Group. 
Differs from the appendieulata group only in the females having bare eyes and 
relatively narrow fronts (index 3 to 4) ; otherwise they closely resemble circumdata 
and its allies. No males are known. 

Dasybasis TRILINEALIS ( Fcrgusou & Henry). 

*Tabanus triUneaJis Ferguson and Henry, 1920. Proc. Linn. Soc. N.S.W., 44, p. 841. 
Type 5, from Kendall, N.S.W., in the Australian Museum, Sydney. 

An ornate species, with sharply defined scutal vittae, and almost confluent median 
abdominal triangles. 

Distribution. — South-eastern Queensland; north coastal New South Wales to Sydney. 

Dasybasis kewensis (Ferguson & Henry). 

*Tabanus keivensis Ferguson and Henry, 1920. Proc. Linn. Soc. N.S.W., 44. p. 843. 
Type 5. from near Kendall, N.S.W., in the Australian Museum, Sydney. 

Resembles a rather dark circumdata. 

Distribution. — South-east Queensland (Nerang) ;. north coastal New South Wales 
to Sydney. 

Dasybasis eidsvoldensis (Taylor). 

*Tabanus eidsvoldensis Taylor, 1919. Proc Linn. Soc. N.S.W., 44, p. 49. Type $, 
from Eidsvold, S.Q., in the School of Public Health and Tropical Medicine, Sydney. 

Extremely like acutipalpis, except for bare eyes and wider antennal plate. 

Distribution. — South-eastern Queensland. 

Dasybasis milsonis (Ricardo). 

*Tabanus milsonis Ricardo, 1917. Ann. Mag. not. Hist.. (8), 19, p. 220. Type J, 
from Milson I., N.S.W., in the British Museum (Natural History). 

Larger (15 mm.) and more robust than the other members of the group, and with 
hairs on the eyes just preceptible at x 15. 

Distribution. — Northern New South Wales, from Narrabri to Hawkesbury R. 

The appendieulata Group. 
Small to fairly large (10-15 mm.), usually dark, compact, hairy species, with fairly 
well-defined scutal and abdominal patterns. Eyes densely hairy in both sexes; varying 
in males from unenlarged, with entirely small facets, to somewhat swollen, with 
moderately enlarged upper facets. Females with frons very wide to medium (index 
1-5 to 3, rarely 3-5), markedly diverging to nearly parallel; callus nearly always full 



BY I. M. MACTvERRAS. 179 

width of frons at base, often with median extension; face usually truncate, and 
separated from paratacials by a clearly defined suture on each side (cf. viacrophthalma 
group); terminal segments of abdomen dorsoventrally compressed (Text-fig. 3). 

This group is essentially southern, not being known north of Eidsvold in south 
Queensland. Its nearest relatives are in southern South America and South Africa. 
Hardy (1948) separated the first four species, with very wide, strongly diverging frons, 
as the subgenus Dasybasis sens, strict., from the remainder, which he included in the 
subgenus Dolichaplia Enderlein. It seems to me that the respective genotypes merely 
represent one end and about the middle of a continuous series in which any division 
is purely arbitrary. The group, as now defined, is sufficiently compact to stand as a 
single unit. 

Dasybasis appendiculata Macquart. 

*Dasybasis apiiendiculata Macquart, 1847. Dipt, exot., Suppl. 2, p. 25. Type c?J, 
from New Holland, in the British Museum (Natural History). The J" bears Macquart's 
"n. gen., n. sp." label, and must be regarded as the holotype. 

A small, hairy, blackish species; with very wide, strongly diverging frons (index 
1-5 to 1-8); transverse callus, with short median extension; slender, blackish antennal 
plate, which is gently curved dorsally; and spotted wings; the only member of the 
group with the eyes banded in life (cf. Cydistoviyia musgravii) . The antennal style 
usually has four annuli, but they are often obscure, and it was described as three- 
annulate by Macquart, which led Enderlein (1922) to include Dasybasis in the 
Haematopotini. 

Females can be divided into two, fairly distinct variants. One, represented by 
the type female and two recent females from Victoria, is smaller (9-10 mm.), with 
almost the whole frons and subcallus shiny, and extensive shiny black areas on the 
parafacials and face. The other, much commoner, usually somewhat larger (10-12 mm.) 
race has the shiny areas limited to the callus and variably on the upper part of the 
subcallus, but the parafacials and face entirely tomentose; it tends to merge with the 
still larger (12-13 mm.) froggatti in southern New South Wales. Males* have the 
parafacials and face tomentose, and can be distinguished only by size, slight differences 
in coloration and general appearance, and association with the females. The type male 
agrees best with males that were taken in company with the two Victorian females 
mentioned above, and it is to the rarer, southern race, therefore, that the name should 
be strictly applied. 

Distribution. — New South Wales, from Mudgee to southern highlands (common 
race); Victoria (boih races). 

Dasybasis froggatti (Ricardo). 

*Tabaniis froggatti Ricardo, 1915. Ann. Mag. nat. Hist., (8), 16, p. 285. Type J, 
from S. coast of N.S.W., in the British Museum (Natural History). 

Distinguished from the commoner race of appendiculata only by generally larger 
size, more robust build, somewhat narrower frons (index 2), and more extensively 
darkened tibiae. There are intermediates, particularly at Canberra, and I doubt that 
specific separation can be maintained. 

Distribution. — New South Wales (Lismore, south coast, southern highlands); the 
only species that is abundant above the tree-line on Mt. Kosciusko. 

Dasybasis gentilis (Erichson). 

Tabanus gentilis Erichson, 1842. Arch. Naturges., (1), 8, p. 271. Type J, from 
Tasmania, in the Berlin Museum, compared with specimens from Australia by Mr. H. 
Oldroyd. 

Tabanus gentilis imminutus Hardy, 1948. Proc. R. Soc. Qd., 58, p. 171. No type nor 
definite type locality indicated. 

A brightly marked, brown species, with wide frons (index 1-8 to 2) relatively wide, 
red-brown antennal plate, brown legs, and conspicuously spotted wings. The antennal 
character relied on by Hardy to distinguish the mainland race varies considerably in 
both mainland and Tasmanian specimens. 



180 ANNOTATICU CATALOGUE OF DESCRIBED AUSTEAI>IAK TABANINAE, 

Distribution. — Mountains of New South Wales, from the Dorrigo plateau to Kiandra; 
Victoria: Tasmania. 

Dasybasis neolatifkons (Ferguson & Hill). 

*Tabanus latifrons Ferguson, 1921. Proc. R. Soc. Vic, 33, p. 19. Type $(j', from 
Cradle Mt., Tasmania, in the Australian Museum, Sydney. Not T. latifrons Zetterstedt, 
1842, Palaearctic. 

Tabanus neolatifrons Ferguson and Hill, 1922. Pkoc. Lixiw Soc. N.S.W., 47, p. 264. 
Nom. nov. 

A distinctive, medium-sized (12-13 mm.), black species, with frons index 2 to 2-5, 
antennal plate moderately wide and rounded dorsally, grey, unspotted wings, and small 
median white spots on abdominal tergites. 

Dist7-ibution. — Flinders I.; Tasmania. 

Dasybasis imperfecta (Walker). 

*Tabanus imperfectus Walker, 1848. List Dipt. Brit. Mus., 1, p. 179. Type J, from 
New Holland, in the British Museum (Natural History). The label "New South Wales, 
ex col. Saunders" is a later addition. 

A small (10 mm.), slender, black species, with greyish-white hairs on the eyes (a 
distinguishing feature of this and the next species), medium, diverging frons (index 
2-5 to 3), slender, gently curved antennal plate, and narrow pale apical bands on 
abdominal tergites. It was correctly identified by White (1915) and Ferguson (1921a, 
1921&), but not by Hardy (1952). 

Distribution. — Tasmania. 

Dasybasis edentula (Macquart). 

*Tabanus edentulus Macquart, 1846. Dipt, exot., Suppl. 1, p. 34. Type J. from 
Tasmania, in the Paris Museum. 

*Tabanus antecedens Walker, 1854. List Dipt. Brit. Mus.. 5, Suppl. 1, p. 253. 
Type 2, from Tasmania, in the British Museum (Natural History). Not T. antecedens 
Walker, 1848, J' (see under gregaria). 

A black species, with white hairs on eyes, and narrow, pale bands on abdominal 
tergites. Distinguished from imperfecta by larger size (12-14 mm.), more parallel-sided 
frons, usually somewhat different callus, and wider antennal plate. It is quite easy to 
separate the two series, even with a hand-lens (,x 8 ) , and I have no doubt that they 
are specifically distinct. 

Macquart's name has been used by different authors for two different species. This 
proved to he due to the fact that, alone among his species that I have studied, he 
attached "n. sp." labels to two specimens. One, in the Paris Museum, agrees with his 
description and his statement that the specimen was in "Museum", and I select it as 
lectotype in order to remove any doubt. It is now headless, but it is accompanied by 
a perfect specimen with the same collector's or accession number ("73 44"), and both 
are undoubtedly the species at present under consideration. This type is evidently the 
specimen referred to by Ferguson (1921?)), when he suggested the synonymy accepted 
here. White (1915) and Ricardo (19156) correctly identified the antecedens of 1854, 
but not edentula. Hardy (1934, 1939, not 1948) accepted Ferguson's suggested 
identification. 

The second "type" specimen is in the British Museum, and it does not agree with 
Macquart's description. It is from the Bigot collection, is labelled "Van Dieman", and 
is a medium-sized (12-5 mm.), smooth, brown species, identical, I believe, with 
Tasmanian specimens of acutipalpis. It provided the basis for the name "edentulus" 
being attached to the common, smaller, brown Tasmanian species by White (1915), 
Pdcardo (1915&), Taylor (1917&, 1918), Ferguson and Henry (1920), and Ferguson 
(1921a, 1921& in part). 

Finally, Hardy (1948) synonymized antecedens Walker, 1854, with flindersi Ferg., 
which it does not at all resemble. 

Distribution. — Tasmania. 



BY I. M. MACKEREAS. 181 

Dasybasis gejiella (Walker). 

*Pangonia geviella Walker, 1848. List Dipt. Brit. Mus.. 1, p. 1.39. Type 5' from 
Western Australia, in the British Museum (Natural History). 

This species had not been recognized, except that Ferguson had left an ms. note 
that it was a tabanine with hairy eyes, and there are two females correctly identified 
in the Australian Museum. It proved to be a not uncommon, western species, with a 
broad, strongly angled antennal plate, like gregaria. but with a wide frons (index 1-8 
to 2) and the wings spotted as in ajjpendiculata. 

Distribution. — South-western Western Australia. 

Dasybasis gregaria (Erichson). 

Tabanus gregarius Erichson, 1842. Arch. Naturges.. (1), 8, p. 271. Type 5- from 
Tasmania, in the Berlin Museum; examined by Mr. Oldroyd, and a specimen returned 
by him as agreeing with it. 

*Tabanus antecedens Walker, 1848. List Dipt. Brit. Mus., 1, p. 178. Type c?, from 
New Holland, in the British Museum (Natural History). Not T. antecedens Walker, 
1854 (see under edentula) . 

*Tabanus flindersi Ferguson, 1921. Rec. 8. Aust. Mus., 1, p. 374. Type 5. from 
Flinders I., Bass Strait, in the South Australian Museum, Adelaide. 

A medium-sized (12-13 mm.), strongly built, dark brown species, characterized by 
a broad, blackish antennal plate with prominent, rectangular dorsal angle; frons index 
about 2-5; wings suffused with brown anteriorly and variably along the veins in southern 
specimens, more or less uniformly greyish in northern ones. 

The identity of gregaria had been obscure, until Mr. Oldroyd returned a typical 
southern flindersi as agreeing with the type. Walker's 1848 type of antecedens has lost 
the antennae; it is dark, thickset, with the upper facets of the eyes moderately enlarged 
but not sharply differentiated from the lower, and the hairs on face and parafacials 
predominantly brown. It is difficult to identify precisely, but it agrees distinctly better 
with males of gregaria from coastal New South Wales than with males of indefinita 
from the same area. 

Distribution. — Coastal New South Wales, from Woy Woy to Jervis Bay; Tasmania; 
possibly also South Australia. 

Dasybasis moeetonensis (Ferguson & Hill). 

*Tabanus confusus Taylor, 1917. Proc. Lints^. Soc. N.S.W., 42, p. 523. Type J. from 
Brisbane, S.Q., in the Queensland Museum, Brisbane. Not T. confusus Walker, 1848, 
Nearctic. 

Tabanus moretonensis Ferguson and Hill, 1922. Proc. Linn. Soc. N.S.W., 47, p. 264. 
Nom. nov. 

The northern representative of gregaria, from which it differs in being smaller, 
smoother, with clearer wings, and the upper facets of the eyes of the males larger and 
more definitely marked off from the lower. It might be better treated as a subspecies 
than as a full species. 

Distribution. — South-eastern Queensland. 

Dasybasis dtjbiosa (Ricardo). 

*Tabanus dubiosa Ricardo, 1915. Ann. Mag. not. Hist., (8), 16, p. 284. Type $, from 
Burnett R., S.Q., in the British Museum (Natural History). 

A small-medium (10-13 mm.), compact, smooth, black species; frons index 3 to 3-5; 
antennal plate wide, with well-defined dorsal angle: wings clear; pale abdominal 
pattern well defined, with large median triangles on the tergites. 

Distribution. — South-eastern Queensland. 

Dasybasis dtjbiosa indefinita (Taylor). 
*Tabanus indefinitus Taylor, 1918. Rec. Aust. Mus., 12, p. 68. Type J, from Nepean 
R., N.S.W., in the .Australian .Museum, Sydney. 



182 ANNOTATED CATALOCUE OF DESCRIBED AUSTRALIAN TABANINAE. 

Move hairy and more thickset tlian the nortliern race, usually more brown in 
general colour, and with smaller median and lateral triangles on the abdominal tergites. 
Hardy (1952) sank both duMosa and indeflnita under imperfecta, which he misidentilied. 
As Queensland specimens can be distinguished even from the darkest ones from New 
South Wales, I feel that subspeciflc separation of the first two is justified, with imperfecta 
quite distinct. 

Distribution. — South Queensland (Stanthorpe, Eukey) ; New South Wales, from 
the Dorrigo plateau to Canberra, and on the coast south of Sydney. 

Dasybasis hobartiensis (White). 

*Tabanus hohariiensis White, 1915. Pap. Proc. R. 8oc. Tas. for 191.5, p. 13. Type $, 
from Hobart, Tasmania, in the British Museum (Natural History). 

A small (11 mm.), compact, brown species; apparently the Tasmanian repre- 
sentative of indeflnita. from which it is distinguished by smaller size, paler hairs on 
eyes, somewhat wider frons (index 2-5), and antennal plate with rounded dorsal angle. 
Possibly it should be treated as a third subspecies in the complex. 

Distribution. — Tasmania. 

Dasybasis tasmaniensis (White). 

*Tabaniis tasmanien.iis White, 1915. Pap. Proc. R. Soc. Tas. for 1915, p. 8. Type 5)^, 
from Bagdad, Tasmania, in the British Museum (Natural History). 

This species has not been identified correctly in Australia, Ferguson (1921?)) giving 
the name to a small, black species close to (or a variant of) imperfecta, and Hardy 
(1934) treating it as a synonym of gregaria^ It is apparently rare, because I brought 
the paratype female (from the same place and date as the type) back with me, and 
have not been able to match it exactly among 394 specimens of Tasmanian Dasybasis 
available for review. It is very close to hobartiensis. differing only in larger size 
(12 mm.), narrower frons (index 3), somewhat differently shaped callus, which is 
slightly narrower than the frons at the base (Text-fig. 9), narrower antennal plate, and 
brown suffusion of the wings anteriorly and narrowly along the veins. However, the 
two specimens of tasmaniensis do stand out when placed alongside specimens of 
Itobartiensis. and I think that they probably represent a distinct species. 

Distribution. — Tasmania. 

Dasybasis circumdata (Walker). 

*Tabanus circumdatus Walker, 1848. List Dipt. Brit. Miis.. 1, p. 185. Type 5, from 
unknown country, in the British Museum (Natural History). 

*Tabanus fraterculus Macquart, 1850. Dipt. exot.. Suppl. 4, p. 30. Type J, from 
"Tasmanie" (erroneous), in the Paris Museum. 

*Tabanus brevidentatus Macquart, 1855. Dipt. exot.. Suppl. 5, p. 28. Type $, from 
Sydney, N. Holland, in the British Museum (Natural History). 

All three types are in poor condition (that of circumdata has had the head gummed 
on, and has lost The third antennal segment and the apical four segments of the 
abdomen ) , but they seem to me to agree better with the larger, common, eastern main- 
land species, than with either of the others that resemble it. In spite of White's and 
Austen's earlier determinations, I cannot associate circumdata satisfactorily with the 
species now identified as hebes. 

This species is about the same size as hebes, but is distinguished from it by usually 
darker colour (melanic forms are frequent, especially in the mountains), more truncate 
face, better defined facial sutures, and especially by the shape of the eighth sternite 
(Text-flg. 5). Since returning to Australia, I have found that the terminal segments 
can be revealed without appreciable damage to the specimen, by simply chipping away 
the free parts of the seventh tergite and sternite with a fine pin, working inwards from 
each side to avoid risk of damaging the underlying structures. 

Distribution. — Eastern New South Wales; Victoria; with possible extensions to 
south Queensland, South Australia and Western Australia; not known from Tasmania. 



BY 1. M. MACKERRAS. 183 

Dasybasis acutipalpis (Macquart). 

*Tabanus acutipalpis Macquart, 1838. Dipt, exot., 1. (i), p. 131. Type $, from 
King I., "Oceanie" (at least latter erroneous), in the Paris Museum. 

*Tabanus abstersus Walker, 1850. Insecta Saimd. Dipt., 1, p. 58. Type $, from 
New South Wales, in the British Museum (Natural History). 

Tabanus edentulus White, 1915, and other authors, nee Macquart (see under 
edentula). 

There are three females in Paris identified by Macquart, the type being labelled 
"Nouv. Holl., Durville" (the description gives "De I'lle King, dans I'Oceanie. M. 
Durville. Museum"), and there is also a female "cotype" in the British Museum labelled 
"Tasmanie". All four seemed to be conspeciflc, but I had difficulty in identifying them, 
until Professor Seguy very kindly sent me the series from Paris for more detailed study. 
The type proved to have a wide eighth sternite, like Text-figure 5, and to agree also in 
external characters with the smaller of the two, common, brown Tasmanian species, 
although the antennal plate was a little wider than is shown in Text-figure 8. 

Mainland specimens, to which the name abstersa strictly applies, may represent a 
distinct race, and are usually distinguishable from eircumdata by smaller size (10-12 
mm. ) , slightly different frons, and narrower antennal plate, with more rounded dorsal 
angle (Text-fig. 7); melanism is equally common. I am inclined, at present, to think 
that eircumdata and acutipalpis are separate, variable species, with overlapping pheno- 
types, but they may completely fuse. 

Distribution. — All States, from south-eastern Queensland to Tasmania and south- 
western Western Australia. 

Dasybasis innotata (Ferguson & Henry). 

*Tabanus innotatus Ferguson and Henry, 1920. Proc. Linn. Soc. N.S.W., 44, p. 847. 
Type 5, from Kendall, N.S.W., in the Australian Museum, Sydney. 

A distinctive, slaty black, long-bodied (14-15 mm.) species, with moderately wide 
frons (index 2-5 to 3), blackish antennae, grey pleura, black legs, quite dark grey 
v/ings, and a row of small, white, median spots on the abdominal tergites. 

Distribution. — Coastal New South Wales, from Kendall to Moruya. There is a 
closely related species (or subspecies) in Western Australia. 

Dasybasis milsomensis (Ferguson & Hill). 

*Tabanus milsoni Taylor, 1917 (Apr.). Proc. Linn. Soc. N.S.W., 41, p. 760. Type $, 
from Milson I., N.S.W., in the School of Public Health and Tropical Medicine, Sydney. 
Not T. milsonis Ricardo, 1917 (Feb.), from same type locality (see under trilinealis 
group). 

Tabanus milsoniensis Ferguson and Hill, 1922. Proc. Linn. Soc. N.S.W., 47, p. 265. 
Nom. nov. 

A small (10 mm.), obscure, greyish-brown species, with medium frons (index 8). 
short, wide, strongly angulate, orange-brown antennal plate, and well-defined pale median 
triangles on the abdominal tergites. 

Distribution. — Coastal New South Wales, from Milson I. to Sydney. 

There are also three small, greyish, undescribed species in this group. One is 
represented by the mislabelled "type" of neocirrus (Ric. ) and other specimens from 
South and Western Australia; the second is from the same States; and the third is 
from western New South Wales. 

Unidentified Species. 
The types of the following species cannot be traced, and no one has been able to 
identify them satisfactorily from the descriptions. It may be noted that none of the 
types of Macquart's 1846 species which were recorded as from the collections of the 
Marquis Spinola and of Guerin and Reiche have been found in Paris or London. These 
include Pangonia singularis, Pangonia bicolor and Tabanus pallipennis, as well as those 
listed below. 



184 ANNOTATED CATALOGUE OF DESCRIBED AUSTRALIAN TABANINAE, 

Tahanus hrevivilta Walker, 1848. List Dipt. Brit. Mus.. 1, p. 157. c?. "New Holland. 
From Mr. Gould's collection." 

Tahanus funebris Macauart, 1846. Dipt, exot., Suppl. 1, p. 33. $. "De la Nouvelle- 
Hollande. MM. Reiche et Guerin." 

Tahanus nigriventris Macquart, 1846. Di2)t. exot.. Suppl. 1, p. 34. J. "De File Sidney. 
Collection de M. le Marquis Spinola." 

Tahanus propinquus Macquart, 1855. Dipt, exot., Suppl. 5, p. 27. J. "De la Nouvelle- 
Hollande. Sydney. M. Bigot." The type should be in the British Museum, but there is 
only a specimen labelled "propinquus c?" in Bigot's handwriting. It has lost its head 
find half the abdomen, and Mr. Oldroyd thinks that "what is left of it looks uncommonly 
like an African T. ustus." 

Tahanus remotus Walker, 1848. List Dipt. Brit. Mus.. 1, p. 177. $. "New Holland. 
Presented by the Haslar Hospital." 

Tahanus sidneyensis Macquart, 1846. Dipt, exot., Suppl. 1, p. 33. "De File Sidney. 
Collection de M. le Marquis Spinola." 

Excluded Species. 

Three, previously unrecognized, Australian species belong to other subfamilies. 

Tahanus bifasciatus Macquart, 1834. Hist. nat. Ins., Dipt., 1, p. 201. Believed to be 
a synonym of Scaptia (Scaptia) guttata (Don.), Pangoniinae. 

*Tahanus erraticus Walker, 1848. List Dipt. Brit. Mus., 1, p. 189. Is a species of 
Ectenopsis {Ecteno'pais) , Pangoniinae, from Western Australia; does not displace any 
currently used name. 

*Tahanus lunulatus Bigot, 1892. Mem. Soc. zooL Fr., 5, p. 688. Is a species of 
Mesomyia (Vepriella), Chrysopinae, replacing *insularis (Ric). 

The following species are excluded from the Australian list. 

*Atylotus claripennis Bigot, 1892. Mem. Soc. zool. Fr., 5, p. 675. Type J, from 
"Australie", in the British Museum (Natural History). Neotropical; replaces Tahanus 
hookeri Knab (Fairchild, 1956). 

*Holcopsis fenestrates Enderlein, 1925. Mitt. zool. Mus. Berl.. 11, p. 372. Type 2, 
doubtfully from "S.W. Australia", in the Berlin Museum. Unlike any known element in 
Australian fauna; possibly Neotropical. 

*Stibasoma Uemiptera Surcouf, 1912. Bull. Mus. Hist. nat. Paris, ann. 1912, (2), 
p. 62. Type 5- from "Nov. Holl.", ex col. Bigot, in the Paris Museum. Neotropical; 
probably Dasychela (Fairchild, 1956). 

*Tabanus leucophilus Walker, 1848. List Dipt. Brit. Mus., 1, p. 154. Type $• fi'om 
"New Holland", in the British Museum (Natural History). Belongs to a distinctive 
group of southern Palaearctic (desert) species of Tahanus, but could not be" matched 
exactly in the British Museum collection. 

*Tahanus limhatinevris Macquart, 1847. Dipt, exot., Suppl. 2, p. 16. Type J, "De la 
Tasmanie. M. Bigot", in the British Museum (Natural History). (Not T. limhatinevris 
Macquart, 1850 — see under Cydistomyia alternata.) A synonym of the Nearctic Tahanus 
ahdomhialis Fabricius, by comparison with named specimens; synonymy suggested, and 
concurred in, by Dr. C. B. Philip. 

*Tahanus ruhricallosus Ricardo, 1914, New Caledonia. Recorded by Hardy (1948) 
from New South Wales; misidentification of Dasybasis macrophthalma (Schiner). 

Tahanus serus Walker, 1862, Cei'am. Recorded by Summers (1912) from Northern 
Territory; misidentification of Tahanus dorsohimaculatus Macquart. 

"■Tahanus similis Macquart, 1850. Dixit, exot., Suppl. 4, p. 31. Type 2. "De la 
Tasmanie. M. Bigot", in the British Museum (Natural History). Agrees with the 
Nearctic *Tahanus lineola scutellaris Walker, by comparison of types by Dr. C. B. Philip 
and I.M.M. 

Tahanus spoliatus Walker, 1860, Celebes. Taylor (1918) noted that Miss Ricardo 
had identified a male in the Australian Museum as T. spoliatiis. Hardy (1948) pointed 
out that the specimen was really Tahanus parvicallosus Ricardo. 



BY I. M. MACKEEKAS. 185 

References. 
Austen, E. E., 1914. — On certain recently described Australian species of Tabanus. Ann. Mag. 

nat. Hist., (8), IS: 263-6. 
EndbrlbiNj G., 1922. — Ein neues Tabanidensystem. Mitt. eool. Mus. Berl, 10: 333-51. 
Fairchild_. G. B., 1956. — Synonymical notes on Neotropical flies of the family Tabanidae 

(Diptera). Smithson. Misc. Coll., 131, No. 3, 38 pp. 
Ferguson^ E. W., 1920. — Entomological notes, (b). Tabanidae (March flies). Rep. Dir. Puhl. 

Hlth. N.S.W. for 1918. Sect. IV, pp. 131-2. 
, 1921a. — A list of the Tabanidae (Diptera) in the South Australian Museum, with 

descriptions of new species. Rec. S. Aust. Mus., 1: 365-79. 
, 1921&. — New Australian Tabanidae, with notes on previously described species. Proc. 

Roy. Soc. Vic, 33 : 1-29. 
FkkgusoNj E. W., and Henry, M., 1920. — Tabanidae from Camden Haven district. New South 

AVales, with descriptions of new species. Proc. Linn. Soc. N.S.W., 44: 828-49. 
Ferguson, E. 'W., and Hill, G. F., 1920. — Notes on Australian Tabanidae. Proc. Linn. Soc. 

N.S.W., 45 : 460-7. 
Ferguson, E. AV., and Hill, G. F., 1922. — Notes on Australian Tabanidae. Part ii. Proc. Linn. 

Soc. N.S.W., 47 : 245-65. 
Hakdt, G. H., 1934. — The genus Tabanus in Tasmania. Stylops, 3 : 43-8. 
, 1939. — Miscellaneous notes on Australian Diptera. V. On eye-coloration, and other 

notes. Proc. Linn. Soc. N.S.W., 64 : 34-50. 
— , 1944. — Miscellaneous notes on Australian Diptera. X, Distribution, classification 



and the Tabanus posticus-grou-p. Proc. Linn. Soc. N.S.W., 69 : 76-86. 

, 1948. — The genus Tabanus in Australia. Proc. Roy. Soc. Qd., 58 : 169-78. 

, 1952. — Miscellaneous notes on Australian Diptera. XV. Tabanus, Heteropsilopus. 



Proc. Linn. Soc. N.S.W., 76 : 222-5. 
Mackbrras, I. M., 1956. — The Tabanidae (Diptera) of Australia. I. General review. Aust. J. 

Zool., 4: 376-407. 
Oldrotd, H., 1949. — The Diptera of the Territory of New Guinea. XIV. Family Tabanidae. 

Part III. Tabaninae. Proc. Linn. Soc. N.S.W., 73 : 304-61. 
RicardOj G. K., 1914. — Notes on the Tabanidae of the Australian region. Ann. Mag. nat. Hist., 

(8), 14: 387-97. 
, 1915a. — Notes on the Tabanidae of the Australian region, Ann. Mag. nat. Hist., (8), 

15 : 270-91. 
, 19156. — Notes on the Tabanidae of the Australian region. Ann. Mag. nat. I-Iist.. (8), 



16: 259-86. 
, 1917. — New species of Tabanidae from Australia and the Fiji Islands. Ann. Mag. 

nat. Hist., (8), 19: 207-24. 
Summers, S. L. M., 1912. — Entomological notes from the London School of Tropical iMedicine. 

No. IV. Blood-sucking Diptera from Port Darwin, Australia. Ann. Mag. nat. Hist., (8), 

10: 222-8. 
SuRCOUF, J. M. R., 1921. — Diptera. Fam. Tabanidae. Wytsman's "Genera Insectorum." Fasc. 

175. 182 pp. 
Taylor, F. H., 1913. — Report of the Entomologist. Tabanidae. Rep. Aust. Inst. Trap. Med. for 

1911, pp. 60-70. 

, 1917a. — Australian Tabanidae (Diptera). No. ii. Proc. Linn. Soc. N.S.AV., 41: 746-62. 

, 19176. — Australian Tabanidae (Diptera). No. iii. Proc. Linn. Soc. N.S.W., 42: 513- 

28. 

, 1918. — Studies in Avistralian Tabanidae. Rec. Aust. Mus., 12 : 53-70. 

, 1919. — Australian Tabanidae (Diptera). No. iv. Proc. Linn. Soc. N.S.W., 44: 41-71. 

, 1926. — Notes on Australian Tabanidae (Diptera). Bull. ent. Res., 17: 193-5. 

AVhite, a., 1915. — The Diptera-Brachycera of Tasmania. Part II. Families Tabanidae and 

Therevidae. Pap. Roy. Soc. Tasm. for 1915: 1-59. 



18G 



OBSERVATIONS ON SOME AUSTRALIAN FOREST INSECTS. 

4. Xyleborus truacatus Ekichson 1842 (Coleoptera: Scolytidae) Associated with 

dying eucalyptxts saligna smith (sydney blue-gum ). 

By K. M. Moore, Forestry Commission of New Soutli Wales. 

(Plate vii; three Text-figures.) 

'' [Read 27th May. 1959.] 



Swnunavy. 

Numerous deaths of trees of E. saligna Smith on State Forests and private property, with 
consequent economic loss, have recently caused concern to Forestry ofHcers and landholders. 

Investigations into tlje causes of these deaths and the association of attack by Xyleborus 
truncaUis Erichson with brown staining in the wood of dying trees are described. 

Biology, hosts and distribution of the insect are given and typical damage is described 
and figured. 

Results of assessments of the quantity of commercial timber destroyed in two areas of 
greatest tree-mortality are Included ; the areas are mapped and their locations are given. 

The factors apparently contributing to attack liy X. trunctus and deaths of trees are 
discussed. 



Introduction. 

Eucalyptus saligna Smith (Sydney blue-gum) occurring on State Forests and private 
property in coastal areas of New South Wales was reported as dying during the period 
1949 to 1958, the increasing number of deaths reaching economic significance toward 
the end of that period. Throughout that time, psyllids of the genus 8pondylias2ns 
(Hemiptera: Psyllidae) were occurring in plagues and causing severe defoliation and 
debilitation of several Eucalyptus species. 

Attention was first drawn to this problem on private property in the Gosford-Wyong 
area by Mr. P. C. Hely, at that time District Entomologist, Department of Agriculture, 
Gosford, who reported that many local residents were blaming Manorina melanophrys 
Latham (bell-birds) for the general debilitation of blue-gums in the area. Hely suggested 
that psyllids of the genus Spondyliaspis were more likely to be the cause (unpublished 
report, 1950). An investigation of the food of the bell-birds was made by Campbell and 
Moore (1956). 

During the years 1953 to 1956 single, scattered, dying trees were observed in many 
areas, and the deaths of E. saligna on State Forests and private property were beginning 
to cause concern to Forestry officers and owners of private property. By this time the 
psyllids had attained plague proportions on trees in numerous gullies, on slopes, and 
on the flats, and had caused partial or complete defoliation continually for some years. 

The initial association of Xyleborus triincatus Er. (PI. vii, lb, Ic) with a dark 
brown staining in the timber of a dying tree was observed during May 1954 at Lisarow, 
N.S.W. The foliage of an Angophora intermedia De CandoUe (rough-barked apple) about 
twenty feet in height, was observed to be brown and apparently dead, and a colony 
of the beetle was present in the trunk of the tree at approximately six feet above ground 
level. Dark brown staining of the wood above and below this colony was extensive, and 
no other evidence of the possible cause of death was observed. A. intermeclia is not 
known to be a host of Spondyliaspis spp. 

Selection of Study Areas. 
Investigations to determine the causes of tree mortalities were begun by the writer 
during November 1956 at Ourimbah State Forest No. 290, where an area of forest 

Proceedings of the Linnean Society of New South Wales, 1959, Vol. Ixxxiv, Part 2. 



BY K. M. MOORE. , 187 

approximately two miles in length and Ave to ten chains in width, extending along the 
centre of a north-easterly slope, was severely infested with 8pondylia.si)is spp., and on 
pioperty of Mr. J. Catt of Lisarow. A study of the biology and ecology of the psyllids 
which was made in those areas will form the subject of a separate paper. 

When a wood-boring beetle was invariably found in association with the debilitated 
trees, many other areas of forest were also examined where the psyllidsi were present 
in large numbers and deaths of E. saligna were occurring. 

When selecting areas for intensive examination, those where logging had occurred 
during the previous ten years were regarded as unsatisfactory for the purpose, because 
the healthier, larger and better quality trees suitable for economic utilization had been 
removed. Those remaining were generally deformed or affected by insect attack (mainly 
by Lepidoptera, Coleoptera or Is'optera), decomposition of the wood where mechanical 
injury had occurred, or fungi. 

The numerous effects of fire in a stand of timber are not completely understood, 
and because these may have introduced unknown variables during investigations, areas 
where fire had occurred within the previous ten years were also considered unsuitable. 

A gully was selected at Lisarow on the opposite side of the creek to the area assessed 
and shown in Text-figure 1, where a stand of large and previously vigorous E. saligna 
trees occurred and where there had been no logging or fire after 1945. The gully 
extended approximately north and south, with a northerly aspect, the timber-stand 
composition on its slopes being principally of E. pilularis J. E. Smith (blackbutt), E. 
jjaniculata J. E. Smith (grey iron-bark), Syncarpia laurifolia Tenore (turpentine), A. 
intermedia, E. saligna and E. acmenioides Schauer (white mahogany), with the two 
latter species predominating. From the central area of the slopes and extending to the 
creek, palms, brush-woods and vines formed a moderate to dense cover beneath these 
species. This area had been severely attacked by Spondyliaspis sp. for at least eight 
years prior to these investigations. 

Methods of Study. 

Twelve trees of E. saligna which appeared almost dead were selected for observa- 
tions on the association of their probable deaths with some cause other than the psyliid 
attack. The twelve trees examined varied in height from thirty to ninety feet, and in 
diameter at breast-height-over-bark from seven to fifteen inches. The trunks were 
apparently free from injury or insect attack, and the trees had previously made 
vigorous, straight growth. Each tree was felled and cut into various lengths, which were 
then examined for evidence of any agent which may have contributed to the tree's 
debility. Attack by Spondyliaspis sp. and X. truncatus was found to occur on each tree. 

There does not appear to be a precise definition of a "dying", "almost dead" or 
"dead" tree, and it was necessary to establish a practical classification of tree condition 
for the purpose of this study. This was formulated by the writer after two years of 
observations on most of the Gosford-Wyong area, which included approximately 150 
separate areas where the psyllids occurred in large populations and trees of E. saligna 
were in various stages of debilitation. From this survey it was assumed that the 
majority of trees bearing dead branches, and with relatively few epicormics, would die, 
and this was considered to be a practical basis for these investigations. 

Results of Investigations and Description of Damage. 

In all areas examined where X. truncatus occurred and damage by the psyllids was 
severe, trees were found to be in various stages of debility. Some showing vigorous 
regrowth of crowns appeared healthy; others carried one or more dead branches with 
variable amounts of foliage or some epicormics on the trunk or branches, while other 
trees had died. 

Many trees severely affected by the psyllids were attacked by X. trimcatus, and in 
these, variable degrees of dark brown staining of the timber occurred contiguous to the 
hole made by the beetle. This staining extended in a strip of variable width, from the 



188 OKSEKVATIONS ON SOME AUSTRALIAN FOREST INSECTS, 

entrance hole and along the sapwood for varying distances toward both the bases and 
the crowns of the trees. In a number of instances it was found to have penetrated the 
truewood. 

Early stages of attack were denoted by a dark brown stain approximately one- 
quarter of an inch in width on the surface of the sapwood, extending above and belov/ 
the hole for from two to six inches. In the later stages, this stain had spread along the 
sapwood and truewood, principally below the holes, becoming paler brown in colour as 
it extended down the trunk, and the bark covering the stained area had eventually 
cracked and died. 

In trees assumed to be almost dead, the stain covered by the dead and dying bark 
sometimes reached to ground-level. The earlier stages showed little or no cracked 
bark. The cracking of the bark always began contiguous to and beneath the scolytid 
hole, spreading downwards. 

In some of the less debilitated trees attack by X. truncatus was not always found. 
Attack in a tree was usually indicated by the presence of one or more dead lower 
branches, with little or no crown foliage, or with epicormics on the trunk or branches. 

During the preliminary investigation of severely debilitated trees, extensive brown 
staining of the wood tissues was associated with scolytid attack in four areas of forest 
which were in some instances as far as twelve miles apart. Live or dead beetles of 
X. truncatus, or its colony chambers, were found in fourteen of the sixteen trees 
inspected throughout these areas. In the other two trees staining was associated with 
borer holes (probably those of this scolytid), but these did not contain beetles or larvae. 

The association of X. truncatus attack with brown staining and the probable death 
of a tree was consistently established in almost every tree out of a total of approximately 
60 examined thereafter. 

The founding of a colony of X. truncatus is not necessary for staining of the timber 
to become evident, or for the death of a tree; a single hole, sometimes containing a 
dead beetle only, was at times sufficient to produce the condition. Initial attack on a 
tree by X. truncatus was usually found in the dead or dying lower branches of the 
tree-crown, and occurred either high or low in the trunk only when a tree was 
apparently considerably debilitated. 

Trees on which the crown or epicormic growth appeared vigorous and healthy as 
though recovering from psyllid attack, and those considered nearest to death, showed 
attack by X. truncatus. Some trees were apparently able to withstand a single attack, 
although others showing a single attack had died. 

Attack by X. truncatus with consequent death of trees sometimes occurs where the 
payllids have played no part in their debilitation, although prolonged attack by 
Spondyliaspis sp., or damage by other agencies, apparently reduces the vitality of trees, 
causing them to become susceptible to attack by X. truncatus. 

Other Species of Wood-bokxng Coleoptera. 

Additional wood-boring beetles observed attacking debilitated E. saligna are as 
follows: 

Anobiidae: Deroptilinus granicollis Lea. 

Bostrychidae: Xylion cylindricus MacL, X. collaris Er., Xylopsocus gibbicolUs 
Macl. 

Brenthidae: Cyphagogus bipunctatus Senna. 

Platypodidae : Platypus australis Froggatt. 

Scolytidae: Xyleborus compressus Lea, X. pseudoangustatus Schedl, X. solidus 
Eichh. 

In the trees examined, attack by X. truncatus always preceded any evidence of 
attack by these other species. 



BY K. M. MOORE. 



189 



Loss OF COMMEKCIAL TIMBER. 

To determine the amount of commercial quality timber represented by the E. salic/na 
killed on two areas of forest, one hundred percent, assessments were made of 2-4 acres 
on the property of Mr. J. Catt at Lisarow (area "A", Text-flg. 1), and 1-05 acres on 
Ourimbah State Forest (area "B", Text-fig. 2). The locations of these areas on the 
Gosford-Norahville military sheet, Zone 8, are shown as insets on the respective figures. 

From the surveyed baselines of each of the assessed areas traverses at 90° and av 
one chain intervals were made. Trees within one half-chain on either side of these 
traverses were assessed, and marked to obviate duplication of assessment. 

On area "A" 92 trees were measured and assessed for the quantity of commercial 
timber which they represented. Fifty-one trees (55%) were classified as dead (of this 
number 19-6% were assessed as "expected to die" because of the extent of their debility 



AREA A 




CONTOURb ON 
THWtRSES 



LtGENO 
DtftD Lsalii 



Text-fig. 1. — Area "A" of assessed E. saligiia, on property of Mr. J. Catt, Lisarow, N.S.W. 
Location 440,600 east, 872,750 north. 



when examined). The amount of commercial timber represented by the fifty-one trees 
was 9,409 super feet hoppus (58% of the total s.f. hoppus of E. saligna on the area), 
and that by the trees classified as "living", 6,813 s.f. hoppus. 

On area "B" 69 trees were assessed. Forty-one trees (59%) were classified as dead 
(this figure included 24% assessed as "expected to die"). The timber represented by 
the forty-one trees totalled 16,156 s.f. hoppus (64-9% of the total s.f. hoppus of E. saligna 
on the area), and that by the trees classified as "living", 8,724 s.f. hoppus. 

Totals for both of the assessments: 

Dead trees: 92 (57%) 25,565 s.f. hoppus. 

Living trees: 69 (43%) 15,537 s.f. hoppus. 
Grand total: 161 trees 41,102 s.f. hoppus. 

Volumes of timber derived from these assessments are based on Forestry Commission 
Recovery Tables for Eucalyptus grandis (flooded gum), Pine Creek S.F. (a closely allied 
species), which were considered adequate, as tables for E. saligna. were not available. 
These figures are indicative of the loss of commercial timber in areas where the 
incidence of mortalities was greatest. ■ 



190 



OBSERVATIONS ON SOME AUSTRALIAN FOREST INSECTS, 



Distribution. 

X. truncatus is widely distributed. The original description by Erichson (1842) is 
based on a specimen from Van Diemen's Land. A specimen from Armidale, N.S.W., was 
described by Lea (1893), who also gave South Australia and Tasmania as localities, 
gpecimens were collected at Forbes and Ourimbah in New South Wales, and also in 
Victoria by Froggatt (1926). Brimblecombe (19.53) records its occurrence in Queens- 
land. It has been collected at Wentworth Falls, Kincumber and Wyong, N.S.W., by the 
writer. 

Hosts. 

Froggatt (1926) recorded E. saligna and E. camaldulensis Dehn. (river red-gum) 
as hosts. E. acnienioides. E. maculata Hook, (spotted gum) and E. citriodora Hook, 
(lemon-scented gum) have been recorded by Brimblecombe (1953). A. intermedia, E. 



hKlh B 





LE5IM 



Text-fig. 2. — Area "B" of assessed E. saligna on Ourimbah State Forest No. 290, Forestry 
District of Newcastle, sub-district of Wyong. Location 434,500 east, 882,075 north. 



Tiropinqua Deane & Maid, (grey-gum) and E. inperita Smith (peppermint) are recorded 
by the writer. 

From the distribution of X. truncatus it is assumed that a wide range of Eucalyptus 
spp. would be attacked. 

Biology. 

Biology of this species is recorded by Froggatt and Brimblecombe. 

Approximately sixty trees of E. saligna in the eight to twenty-flve .years age-group 
were felled in the Gosford-Wyong area during 1957 and 1958 and the following observa- 
tions made. 

X. truncatus has been found only in standing trees. The colony-chambers (PI. vii, 
la) are found in the truewood as deep as two inches within the branches or the trunk, 
the narrowest portion of the chamber being above the entrance tunnel. The colony- 
chambers, which vary in size, are approximately 2 mm. wide, 5 mm. in height, with a 
basal length of 4 mm., their size probably depending on the number of larvae reared in 
them. The walls of the chambers or the holes are sometimes stained a brown to black 



BY K. M. MOORE. 191 

colour which spreads into the sapwood or truewood above and below them. The staining 
may be seen above and below the galleries in the figure. There may be no typical 
chamber such as that figured, but only a more or less horizontal hole from the outside 
of the trunk or limb, and which may contain adults, pupae or larvae. 

One colony contained 13 larvae, most of which were in the last instar, and only 
the base of the colony-chamber was stained. Some live colonies were found in apparently 
dead and dry branches, although it was more usual to And them in timber which still 
contained sap. 

Froggatt and Brimblecombe refer to attack occurring in damaged areas on trees, 
where sapwood had been exposed. In each of the instances observed during these 
investigations, attack occurred through the previously undamaged, smooth bark of E. 
saligna and E. propinqua or the rough bark of A. intermedia. 

The association of X. truncatus with dying trees has not previously been reported, 
although Froggatt recorded that the exposed area of wood surrounding the point of ' 
attack turned brown and died. 

Oviposition apparently occurs over many months in the one colony, for larvae of 
most instars, pupae and adults occur together during most months of the year. During 
these investigations oviposition occurred from October to May, and during the latter 
month a cluster of approximately 15 eggs was found in a colony-chamber containing 
larvae and pupae. This suggests that the one colony-chamber may be utilized for the 
rearing of more than one generation of beetles. 

Beetles may bore into either a limb or tree-trunk and die without founding a 
colony or without evident staining of the timber. Some apparently killed by sap-flow 
or gum-flow were found in the entrance holes. 

The parent beetle appears to attend a colony for some months, and some have been 
found in entrance holes around the external edge of which a white, powdery substance 
(probably the excreta of the larvae, or excess fungal growth) had been removed from 
the gallery. A single exit hole is constructed from the end of the colony-chamber 
opposite to the entrance hole made by the parent when founding the colony. 

The life cycle occupies approximately three months in the warmer weather, and 
six months or more for the overwintering generation. Emergences of adults from a 
colony may occur from September to about May, while pale coloured adults, together 
with last instar lavae, were taken during July. It is presumed that adults would not 
have emerged from the tree until the following spring. 

No parasites of X. truncatus were found during these investigations. 

Adult bees of Hylaeus aralis Ckll. (CoUetidae: Hylaeinae) were reared from gallery- 
chambers. Their larvae overwinter in thin skin-like cells in the galleries. 

Descriptions. 

The adults, 2-5 mm. to 3 mm. in length, may be identified by the truncate elytra, a 
feature which is apparently confined to this one species of the Australian Scolytidae. 
They superficially lesemble species of the Bostrychidae, in which family this feature is 
comparatively common, and Lea (1893) initially placed X. truncatus in that family. 

The taxonomy of X. truncatus has been referred to by Brimblecombe (1953). 

Larva. — There is apparently no previous description of a last instar larva of X. 
trimcatus, although Froggatt figured a larva without designating to it any particular 
instar. Last instar (Text-fig. 3) : Length approximately 3-5 mm. to 38 mm. Head- 
capsule pale cream with mouthparts varying from tan-colour to dark brown; remainder 
of larva opaque white; cylindrical; the exoskeleton more or less covered with micro- 
setae which are visible at high magnifications; apodous; arcuate, with the abdomen 
prominently deflexed from about the fourth abdominal segment, so that the distal 
segments are often at an angle of about 90° to the proximal segments; pseudopods 
present on each thoracic segment; tenth abdominal segment small, and consisting of 
little mere than the anal aperture; the dorsal aspect of each segment except abdominal 



192 



OBSERVATIONS ON SOME AUSTRALIAN FOREST INSECTS, 



segments nine and ten is divided transversely to form an anterior and a posterior 
protuberance approximately equal in length. 

Setal arrangement on the head-capsule and body-segments is shown in Text-figure 3. 

Pupa. — The pupa, about 3 mm. in length, is at first white, later becoming yellow, 
then pale brown prior to the emergence of the adult. 

Discussion. 

The cause of deaths of E. saligna was not determined, and these observations 
present a basis for further investigations. 

Species of the Scolytidae are known to be vectors of Dutch elm disease, with 
consequent deaths of trees, in forests of Europe and America (Collins et al., 1936; 




Text-fig. 3. — Setal arrangement on body-segments and head-capsule of last instar larva 
of X. truncatus. 



Parker et al., 19411, and it is suggested that X. truncatus is similarly associated with 
some pathogen or toxic agent capable of causing tree mortalities. It is most unlikely 
that the limited damage caused by these beetles would be the direct cause of death. 

Certain weather conditions or physiological conditions of trees may be essential 
for optimum effect of the mortality factor which is not necessarily associated with 
psyllid or scolytid attack. 

The considerable number of dead and dying trees occurring in areas of large 
psyllid populations and where X. truncatus attack occurred was generally associated 
with debilitation of trees apparently due to, or accentuated by, persistent psyllid attack. 
However, the A. intermedia attacked by X. truncatus was not debilitated by psyllids, 
which suggests that site-favourability may be a factor contributing to debilitation, and 
thus inducing attack by X. truncatus. This may be the principal factor operating in all 
instances. 



BY K. M. MOOKE. 193 

From an examination of meteorological records supplied by the Narara Citrus 
Experiment Station, it is evident tliat the reports of large populations of psyllids 
together with debilitation of E. saligna correspond with the years of abnormal rainfall 
experienced from 1949 to 1956. The average annual rainfall for that period was 69-29 
inches. Rainfall exceeded 86 inches for each of two years and exceeded 72 inches each 
year for a further three years during those eight years. 

For comparison, the average annual rainfall for the years 1935 to 1941 was 37-16 
inches for the seven-year period, while the average annual rainfall for the intervening 
seven years of 1942 to 1948 (45-06 inches) approached the normal figures for precipita- 
tion on the area, which for the 41 years from 1917 to 1957 averaged 49-83 inches per 
annum. 

At present an analysis of factors contributing to deaths of the trees is hypothetical, 
but the hypothesis considered most tenable is that the abnormal rainfall adversely 
affected the physiology of Eucalyptus and other species generally, making them suscep- 
tible to heavy attack by psyllids. 

It appears that moderate temperatures together with high relative humidity favour a 
large increase in psyllid population in the areas studied (probably through conditions 
unfavourable to their natural parasites), causing the progressive debilitation of the 
trees and culminating in a greater susceptibility to attack by X. truncatus. 

Observations on these factors are being continued. 

Acknowledgements. 

Acknowledgement is made to Mr. K. G. Campbell for assistance given, particularly 
with the timber assessments, and to Mr. P. Hadlington, officers of the Entomological 
Research Section, Forestry Commission of New South Wales. The writer is also grateful 
to Messrs. J. Catt and W. Mann of Lisarow, who readily gave permission to investigate 
and fell trees on their properties, and to many who assisted in the preparation of the 
manuscript. 

References. 
Brimblecombe^ a. R., 1953. — An Annotated List of the Scolytidae Occurring in Australia. Q'd. 

J. Agric. Sci., 10 (3) : 167-205. 
Campbell, K. G., and Moore, K. M., 19 56.- — An Investigation of the Food of the Bell-bird 

Manorina melanophrys Latham, Proc. Roy. Zool. Soc. N.S.W., pp. 72-73. (May 1957.) 
Collins, C. W., et al., 1936. — J. Econ. Ent., 29 (1) : 169-176. 
Erichson, "W. F., 1842. — Arch. Naturgesch., 8 (1) : 212. 
Froggatt, W. W., 1926. — Aust. For. J., 9 : 144-5. 
Lea, a. M., 1893.- — Descriptions of New Species of Bostrychidae. Proc. Linn. Soc. N.S.W., 8 

(2) : 317-323. 

, 1904. — Australian Coleoptera. Proc. Linn. Soc. N.S.W., 29 (1) : 106. 

Parker, K. G., et al., 1941. — Phytopathology, 31 (7) : 657-663. 

EXPLANATION OF PLATE VII. 
la. Colony-chambers of X. truncatus (Scolytidae) in E. saligna with associated staining of 
the timber. 16. Adult beetle of X. truncatus (lateral view), le. Adult beetle of X. truncatus 
(dorsal view). 

Photographs by D. Rose. 



194 



THE GENUS C0N08TYLIS R.BR. 

I. Leaf An'atomy. 

By J. W. Green, Department of Botany, University of New England, Armidale, 

New South Wales. 

(Twenty-five Text-figures.) 

[Read 24th June, 1959.] 



Synopsis. 
Anatomical features of the leaves of selected species of ConostyJis are described and 
illustrated and their systematic and ecological significance discussed briefly. 



Introduction. 
Occurrence and Hahit. 

The genus Conostylis is restricted to a small area in south-western Australia 
where it is a conspicuous member of the Liliiflorae in all but the wettest vegetation 
communities. 

In habit, there is extreme diversity between the different species. All are perennials 
and many have a strong capacity for vegetative reproduction which may be by means 
of stolons, rhizomes or a method of growth referred to by Bentham (1873) as pro- 
liferous branching. The habit may therefore be prostrate and spreading, caespitose 




Text-fig. 1. — Distribution of the genus Conostylis. The occurrence within this area of the 
species dealt with in the present paper is indicated in Table 2. 

or almost shrubby, some plants attaining a height of 50-60 cm. There is nearly always 
a large amount of sclerenchyma in the leaves, stems and roots, and a number of species 
bear harsh spines on the leaf margins. The flowers and flowering scapes of practically 
all species bear a dense, woolly tomentum of branched trichomes. 

Materials and Methods. 

The data presented have been derived from foliar cross sections prepared largely 
from fresh material fixed in the field, although some species have, of necessity, been 
studied from herbarium material; in the latter case leaves were boiled in water for a 
few minutes before being sectioned. Measurements on specimens, part of the leaves 
of which were fixed fresh, while the remainder was thoroughly pressed and dried, 
showed that the maximum shrinkage likely in dried herbarium material was of the 



Proceedings of the Linnean Society of New South Wales, 1959, Vol. Ixxxiv, Part 2. 



BY J. W. GREEN. 195 

order of 10% (linear measurement) compared with fresh specimens. On hoiling for a 
few minutes the original size was restored. 

Difficulties were encountered in attempts to microtome this material, probahly due 
to the extremely fibrous nature of the leaves of most species. Satisfactory hand sections 
were obtained and were stained with safranin and fast green (Johansen, 1940) and 
made permanent. 

Camera lucida drawings were employed throughout. 

General Structure. 

The chief feature of taxonomic importance is the disposition of sclerenchymatous 
tissue in the vicinity of the vascular bundles. By considering, in addition, lignification 
of the epidermal cell walls, presence or absence of tannin-containing cells, occurrence 
of large subepidermal cavities, leaf shape and the occurrence of surface hairs and 
protuberances, it has been possible to identify the twelve species dealt with here. 

A developmental feature of interest lies in the origin of the two surfaces of the 
lamina. Although exhibiting, in cross section, a structure similar to that of an 
isobilateral leaf, such as is found in species of EremopMla or Eucalyptus, the dis- 
position of the conducting tissues in the mature leaf of Conostylis has resulted from 
the ontological fusion of the two halves of the plicate sheathing base (Text-fig. 24). 
Bjj this process the adaxial surface has been highly reduced and the two lateral 
surfaces of the mature part of the leaf are each abaxial. Arber (1925) has described 
this phenomenon in Romulea, Galaxia and some other monocotyledonous genera. 

Because almost the entire leaf surface is abaxial, some difficulty arises in describing 
the orientation of structures within the leaves. For the purpose of this study the 
terms upper or lower are abandoned in referring to the leaf surfaces; instead, structures: 
in the vicinity of either epidermis will be referred to as outer, while by the inner part of 
the leaf will be meant the region near the imaginary line of fusion of the two halves 
of the lamina. 

Histology. 
Epidermis. 

Typical cells are more or less isodiametric, having a diameter of 20-30/i; the 
largest observed were about SOm in diameter. The cell walls are commonly heavily 
lignified and are often tangentially compressed (Text-fig. 12), but sometimes are thin 
and parenchymatous. 

The thickening material appears to be lignocellulose, said to be a "relatively 
infrequent component of epidermal walls" (Linsbauer, quoted by Esau, 1953). Several 
species show stratification in the walls (Text-fig. 14). Pit pairs are frequently seen» 
particularly on the inner tangential and the radial walls. The cuticle is commonly 
]-2/i thick. Stomata are usually neither deeply sunken nor exserted; exceptions are 
mentioned specifically below. 

The determination of stomatal frequency from the examination of leaf-reconstruc- 
tions from cross sections might prove easier than from surface counts which, in this 
genus, are hampered by the presence of hairs and ridges on the leaf surface in some 
species. Measurements of stomatal frequency and size of guard cells are of interest in 
the case of C. setigera, which is thought to be a polyploid derivative of C. setosa 
(Green, 1958). 

Multicellular trichomes arise from the laminar epidermis of some species, while 
in others marginal spines or setae consist of fibrous outgrowths of the epidermis. The 
leaf blade trichomes are of the branched candelabra type (Esau, 1953) (Text-fig. 2), 
while the marginal spines or setae may be minutely branched on the upper side or 
entire. 

Palisade. 

In the present paper the term palisade is used to denote the principal (and usually 
palisade-like) photosynthetic tissue of the mesophyll. The term storage parenchyma 



196 



THE GENUS CONOSTYLIS. I, 



Storage 

Parench^Tna 

CeUs. 


ins 


ITS 




1 











QO 







CD 
CO 


1 



CO 









CO 

1 



3 

O 

OS 

i 


3 

1=1 

1 


<w 


l^ 





m 





lO 


C-] 


iro 








01 


3 




CO 



CO 












iC 


00 
CO 




1 

CO 




CO 




in 


c 

CO 


■3 


iO 




1 











in 

X 



CO 


tJh 










CO 

X 



CO 


c 


Cuticle 

Thickness. 

(ix) 


1 


i 


- 


-- 


04 


'SH 


^1 


CO 


01 


CO 
0] 




CO 


Other Localities. 


0' 


1 


c 

c 



g 
5 






d 


1 
w 

1; 


rt 

S 

Oj 

s 


c3 






00 
g 

Q 
2 1 

C4 ? 




^ BC 

!2; ^ 


Typical Locality. 


c 


e 


£ 



!2; 




si 
fin 

C 


■§ 


■a 
g 

g 

=s 
3 
M 


|2i 


d 

2 


W 






p; 


i 

1 


ca 

§ 
H 

C« 


c 

3 
1 


S 
P4 


Number 
Examined. 


<M 


-¥ 


01 


CO 


- 


01 


-t 


00 


OJ 


ri^ 


- 






1 

6 


f 

6 

CO 


s 

1 

d 


1 

s 


1 

d 

to 


e 

So 

d 


e 
•1 

d 


1 


e 

.0 
d 

d 




'S 

i 




a 

i 

d 

oi 



.3 X 

> a 






60 .£ 



3 S 

o ss 

.S I 

S* 2 



BY J. W. GREEI^^. 197 

is used to refer to the remaining tissue of the mesophyll, exclusive of the vascular 
tissue. It may or may not contain chloroplasts and is only rarely spongy. 

The palisade consists typically of 2 layers of cells about 15 x 7^, occurring directly 
beneath the epidermis on both sides of the leaf. It may be in the form of a continuous 
band around the leaf or it may be interrupted by girder-like masses of sclerenchyma 
surrounding the vascular bundles and extending on the outside to the epidermis. In 
C. aculeata, whose leaves are characteristically of the interrupted type in the mature 
state, inspection of soft vigorous regrowth following a fire revealed an uninterrupted 
palisade. 

In one species, C. juncea, large subepidermal parenchymatous cells have been 
observed within the palisade. These cells are ovoid in both cross section and 
longitudinal section, are much larger than the palisade cells, and have occurred 
regularly in all leaves examined (Text-fig. 18). 

The cells of the palisade are thin-walled and the tissue may either be sharply 
demarcated from or merge indistinctly into the storage parenchyma. 

Storage Parenchyma. 

The storage cells are usually more or less isodiametric with slightly thickened 
walls and the intercellular spaces inconspicuous. Sometimes the storage parenchyma 
appears spongy with small, regular intercellular spaces (Text-flg. 18). 

Chloroplasts are found usually in diminished numbers toward the inside of the 
leaf, but rarely are almost as dense as in the palisade (Text-fig. 4). Cells with darkly- 
staining contents, presumably phlobaphene, an oxidation product of tannins (Johansen, 
1940), occur prominently in the storage parenchyma and sometimes in the palisade 
(Text-fig. 20 and 25). These cells appear pale to dark brown in unstained sections and 
almost black with safranin and fast green; their presence appears characteristic of 
live species (Table 2). 

Vascular Tissue. 

Almost without exception, the most striking feature of the vascular bundles is 
the occurrence of large areas of sclerenchyma associated with the conducting tissues. 
In many cases the conducting elements are completely surrounded by a mestom sheath 
(Schwendener, quoted by Esau, 1953); sometimes the fibres immediately adjacent to 
the phloem have a distinctive appearance (smaller lumen and differential stainability). 

Often the mestom sheath is in turn surrounded by a prominent parenchymatous 
sheath which may be entire or broken where the phloem fibres extend outward to join 
the epidermis (Text-fig. 6 and 12 resp.). 

The cells of the xylem and phloem call for no special comment. Phloem has been 
observed clearly in sections cut from herbarium specimens more than fifty years old, 
its preservation evidently being due to the protection of the fibrous sheath. 

The vascular bundles occur in two rows in the flat leaves, often with larger and 
smaller bundles alternating (Text-fig. 5). As a result of the developmental fusion of 
the sheathing leaf base, the bundles are oriented with the phloem toward the outside. 
In one species, C. aculeata, the larger bundles are united in pairs in the girder-like 
mass of sclerenchyma (Text-fig. 25), while in another, C setigera, three bundles are 
commonly united within a mass of sclerenchyma which does not reach the epidermis 
(Text-fig. 16). 

Specific Anatomical Descriptions. 

In the following descriptions, twelve species have been chosen as being repre- 
sentative of the main types of anatomical structure observed in the genus. The names 
of other species whose anatomy is essentially similar are appended to each description. 
The arrangement is approximately in order of increasing degrees of xeromorphism, as 
inferred from the disposition of the sclerenchyma, presence of tannin cells, leaf 
tomentum and form and other minor characters (Table 1). 

The descriptions are based on single preparations, chosen as typical of the species 
as a result of the examination of leaves from as many localities as possible. In most 



198 



THE GENUS CONOSTYLIS. I, 




Text-fig-. 2-8. — In the low-power diagrams the palisade is shown in black and the vascular 
bundles in outline. 2, C. candicans, a small but otherwise typical example of a candelabra- 
type trichome ; 3-4, C. stylidioides ; 5-6, C. candicans ; 7-8, C. setosa. 



BY J. W. GREEN. 199 

cases leaves from 3 or 4 localities throughout the range were examined in conjunction 
V7ith the particular leaf described below, but in some cases only one specimen was 
available; the details are set out in Table 1. The dimensions of component cells found 
to be most variable are shown, with measurements from a typical preparation. 

1. CoNOSTYLis STYLiDiomES F. Muell. (Text-fig. 3-4.) 

This species is almost alone in the genus in its paucity of the xeromorpliic 
characters so common in most other species, although a loose tomentum can occur on 
the leaf surface in plants from some parts of the range (notably to the west of 
Geraldton). The remarkable absence of lignification is reflected in the flaccid nature 
of the leaves. 

Epidermal cells are thin-walled and isodiametric; the stomata, if anything, aie 
slightly exserted. The palisade merges indistinctly into the storage parenchyma. The 
number of chloroplasts in the storage parenchyma is scarcely less than in the palisade, 
except for an ill-defined zone of central cells which contain no chloroplasts. The 
vascular bundles are scarcely strengthened, the inner sclerenchymatous sheath being 
reduced to very few (5-8) fibres in most cases. The cells of the parenchymatous sheath 
are slightly more regular than those of the storage parenchyma and contain uo 
chloroplasts. C. prolifera Benth., C. racemosa Benth. 

2. CONOSTYLIS SEORSIFLORA F. Muell. 

Similar to the above, except for the presence, in the single specimen examined, of 
lignified walls in the epidermal cells immediately adjacent to the vascular bundles. 
The remainder of the epidermis is parenchymatous. 

3. CONOSTYLIS CANDICANS Eudl. ( Text-fig. 5-6.) 

A prominent anatomical feature of the leaves is the presence of large, multicellular, 
branched trichomes arising from the epidermis. These give the leaf its characteristic 
mealy appearance in surface view. 

Epidermal cell walls are unlignifled and the cells are more or less isodiametric; 
from the epidermis arise numerous candelabra-type trichomes similar to those occurring 
on the scape and perianth of most species (Text-fig. 2). These consist of a central axis 
of 4 longitudinal series of cells; the lower 1-2 layers bear no projections, but the 3-4 
above bear aciculate protuberances in each of 4 directions, approximately at right 
angles to the axis. The phloem is capped by 3-4 layers of fibres which occasionally 
become produced outward to the epidermis, interruping the palisade. A parenchymatous 
outer sheath occurs prominently. C. dealbata Lindl. 

4. CoNOSTYLis SETOSA Lindl. (Text-fig. 7-8.) 

No single outstanding feature characterizes this species. The leaf is more 
rectangular in section than in other species, the palisade is uninterrupted and pairs 
of bundles occasionally may be more or less united by the fusion of their fibrous 
sheaths on the inside. 

The epidermis consists of isodiametric cells having lignified walls and a small 
lumen. The palisade is continuous and clearly demarcated from the storage parenchyma 
whose cells may be more or less spongy, with very few chloroplasts. The vascular 
bundles are each surrounded by a fibrous sheath, in turn surrounded by a paren- 
chymatous sheath, except in the rare cases when the fibrous sheaths of two opposite 
bundles coalesce. 

5. CONOSTYLIS ANDROSTEMMA (Lindl.) F. Muell. (Text-fig. 9-10.) 

Leaves of this species are more or less circular in section with deep, longitudinal 
grooves. By comparing Text-figures 9 and 21 it may be seen that such grooves may 
occur opposite (i.e., on the same radius) or between the major bundles, in different 
species. Further observations will be necessary to confirm the taxonomic value of 
this character, which appears uniform in the species examined. The sole exception 



200 



THE GENUS CONOSTYLIS. I, 




lOO/L 








Text-fig. 9-14. — 9-lfl, (\ andr ostein ma; 11-12, C. petrophiloides ; 13-14, C. vaginata. 



BY J. W. GREEN. 201 

occurs in an undescribed subspecies of C. androstemma. geographically removed from 
the typical subspecies here described and illustrated, in which the grooves of the single 
specimen examined were between the major bundles. Tannin cells occur in the storage 
parenchyma. 

Epidermal cells are anisodlametric, with walls heavily lignified. In some places 
the space normally occupied by one epidermal cell is occupied by two or three, one 
above the other. This condition will be referred to by the term multiple epider77ial cells; 
in no way does it refer to a hypodermal layer, being restricted to the occurrence of 
multiple cells in an otherwise single-layered tissue (Text-fig. 10, 12, 14, 18, etc.). The 
palisade tissue merges into the storage parenchyma whose cells do not appear spongy 
in spite of small regular intercellular spaces. Many storage cells adjacent to the 
palisade may contain deposits of tannin; one section showed 11 tannin-containing cells 
in the outer storage parenchyma. The usual fibrous and parenchymatous sheaths 
surround the bundles, the latter less prominently than in many other species. The 
fibrous sheath of the larger, alternate bundles often penetrates the palisade to join 
the epidermis. 

6. CoNOSTYLis PETEOPHiLOiDEs F. Muell. ex Benth. (Text-fig. 11-12.) 

A particularly heavily lignified epidermis occurs in this species. Striations, multiple 
cells and pits are characteristic. The palisade is practically continuous. 

The epidermal cells are markedly anisodlametric, being greatly compressed 
tangentially. The cell walls are lignified, striated and pitted and the cells contain, as 
a rule, a small compressed lumen. Multiple cells are very common. Stomata are 
slightly exserted. The palisade lies in a practically continuous band and is clearlj' 
demarcated from the adjoining tissues. The storage parenchyma cells do not appear 
spongy and do not contain chloroplasts. The conducting tissues are immediately 
surrounded by several layers of fibres, in turn surrounded by a parenchymatous sheath 
which is rarely broken. 

7. CoNOSTYLis VAGiKATA Eudl. (Text-fig. 13-14.) 

The leaves are more or less circular in section and the palisade is often in a 
continuous band. When the palisade is interrupted, the epidermis is deeply grooved at 
the point of interruption (Text-fig. 13). Epidermal cell walls are heavily lignified, 
striate and pitted, and multiple cells occur infrequently. 

Epidermal cells are anisodiametric with occasional multiple cells. The walls are 
heavily lignified, with prominent striations and pits on the inner tangential and radial 
walls. The palisade forms a very regular band, sharply distinct from the storage 
parenchyma which contains no chloroplasts. The vascular bundles are each surrounded 
by a fibrous and a parenchymatous sheath, the latter not well defined and occasionally 
broken by the fibrous sheath when it is produced outward to the epidermis. 

8. CoNOSTYLis SETiGERA R.Br. (Text-fig. 15-16.) 

A large area of fibres often encloses three vascular bundles and the palisade is 
always continuous. These features produce an appearance in this species (and its 
relatives, C. melanopogon and C. psyllium) quite unlike that in any other species in 
the genus. 

Epidermal cells are small and isodiametric with the walls heavily lignified. The 
palisade is regular and continuous and the storage parenchyma, small in extent, contains 
chloroplasts. Each fibrous bundle sheath encloses, typically, three vascular bundles; the 
sheath is usually one cell thick on the outside of the bundles, but the whole of the 
more or less triangular region within the three bundles consists of a mass of fibres. 
Pit pairs occur in the walls of many of these cells. The individual bundles are smaller 
than in many other species. A prominent parenchymatous sheath surrounds each 
group of three bundles and its fibrous matrix. C. melanopogon Endl., C. psyllium Endl. 



202 



THE GENUS CONOSTYLIS, I, 







20 




Text-fig. 15-20.— 15-16, C. setigera ; 17-18, C. jnncea; 19-20, C. bealiana. 



BY J. W. GREEN. 203 

9. CoNOSTYLis JUNCEA Endl. (Text-flg. 17-18.) 

This species is distinguished by the presence of large subepidermal cells bearing 
a superficial resemblance to substomatal chambers. The storage parenchyma cells have 
intercellular spaces which give this tissue an appearance unlike that of any other 
species examined. 

Epidermal cells are anisodiametric with occasional multiple cells, and the cell 
v/alls are heavily lignified. The palisade contains prominent subepidermal cells which 
are ovoid in both transverse and longitudinal section; however, the reliability of this 
as a taxonomic character cannot be estimated until the exact nature and constancy of 
the cells is known. They occur irregularly around the leaf, 20 being counted in one 
typical section. The palisade for the most part is continuous, but is usually interrupted 
at one or two places by the outward extension of the fibrous bundle sheath; it merges 
into the storage parenchyma, the cells of which resemble somewhat the typical spongy 
mesophyll found in the leaves of many genera, e.g., Uliiim. Small regular intercellular 
spaces may be seen both in section and surface view of the cell walls. The vascular 
bundles are each surrounded by a fibrous sheath which may protrude through the outer, 
enclosing parenchymatous sheath, to join the epidermis. 

C. involucrata Endl. — Note: Flat-leaved forms of C. jimcea, placed under C- 
involucrata by Bentham (1873), cannot be maintained as a separate species, since 
terete and flat leaves occur frequently on a single plant. The anatomical structure 
of the flat leaves retains the spongy parenchyma and the subepidermal cells described 
above, although the general pattern resembles a flat-leaf type such as C. healiana (Text- 
fig. 19-20). 

10. CONOSTYLIS BEALIANA F. Muell. (Text-fig. 19-20.) 

Many prominent tannin-containing cells occur in the storage parenchyma of the 
leaves of this species. The palisade is well defined and interrupted by many of the 
fibrous bundle sheaths which extend to the epidermis. 

Epidermal cells are isodiametric with heavily lignified walls. The palisade consist* 
typically of 2 cell layers which are sharply distinct from the storage parenchyma. The 
storage cells contain no chloroplasts; they are irregular in shape in the only specimens 
examined, but there is no reason to believe that this is due to distortion arising from 
the preparation of the herbarium specimens used, since this appearance was not seen 
in other species where comparison between living and herbarium material was possible. 
The outermost cells of the storage parenchyma are frequently very large and contain 
tannin-like deposits; no chloroplasts occur in this tissue. The vascular bundles are 
surrounded by a fibrous sheath which often penetrates the surrounding parenchymatous 
sheath to join the epidermis. 

11. CoNOSTYLis CARiciNA Liudl. (Text-fig. 21-22.) 

The leaves are characterized by their large cross sectional area, the presence of 
tannin cells in the storage parenchyma and an extremely heavily thickened epidermis. 

Epidermal cells are tangentially compressed, larger than in all other species, with 
heavily lignified walls; multiple cells occur infrequently. Stomata are confined to 
grooves in the epidermis. The palisade is deeper and consists of more layers of cells 
than in the other species; it occurs in a well-defined layer merging into the storage 
parenchyma. Tannin cells occur infrequently in the outer cell layers of the storage 
parenchyma. The vascular bundles are each surrounded by a fibrous sheath which, in 
many bundles, penetrates the parenchymatous sheath, interrupting the palisade, to 
join the epidermis. 

12. CoNOSTYLis ACULEATA R.Br. (Toxt-fig. 23-25.) 

The most prominent feature of the leaves of this species is the occurrence of girder- 
like masses of sclerenchyma, across the whole width of the section, from epidermis to 
epidermis, enclosing usually two vascular bundles. Several layers of heavily lignified 
hypodermal fibres occur at the margins. 



204 



THE GEJTUS CONOSTYLIS. I^ 



Epidermal cells are Isodiametric with heavily lignified walls; the epidermis is 
trequently undulate with grooves occurring between the vascular girders (see discussion 
under C. androstemma) or the epidermis may be more or less flat. At the margins 







25 




24 

Text-fig. 21-25. — 21-22, C. caricina ; 23, Hypodernial fibres in a form of C. aculeata ; 
24-25, C. aculeata, the low power diagram showing- the plicate sheath near the base of the leaf. 

occur several to many layers of hypodermal fibres (Text-fig. 23). It is upon this last 
character that the species C. hromelioicles is based; in this species the thickening is 
sufficiently heavy to be conspicuous exomorphically. The palisade is much dissected 
by the vascular girders; it merges into the storage parenchyma, cells of which contain 
chloroplasts and frequently tannin. The parenchymatous bundle sheath is reduced to 
a single layer of cells on each side of the fibrous girder. The girder structure is 
developed most strongly in certain related species such as C. robusta and C. bracteata; 



BY J. ^^■. GREEN. 



205 



in all cases it represents the coalescence of two fibrous bundle sheaths when the bundles 
are heavily strengthened and occur directly opposite one another. C. breviscapa R.Br., 
C. bracteata Lindl., C bromelioides Endl., C. preissii Endl., C harperiana W. V. Fitzg., 
C. teretiuscula F. Muell., C. spinuligera F. Muell. ex Benth., C. serrulata R.Br., C 
laxiflora Benth., C. cymosa F. Muell. ex Benth., C. jjJiathyrantha Diels. 



Discussion. 
Ecology and Geographical Distribution. 

Attempts to correlate anatomical features of the leaves of Conostylis species with 
geographical and ecological ranges have proved disappointing. Table 2 shows that 
when the species are arranged according to the degree of development of xeromorphic 
characters the leaf anatomy does not reflect the influence of the present environment. 
Of the two species which never occur on sand (C setosa and C. caricina) no characters 

Table 2. 







£■ 

IS 
■Ms 




.S c 


QJ CD 




* 

>> 

*o 
rr. 


.2 


Annual 
Average 
Rainfall 
(cm.). 


1. c. 


stylidioides 




— 


_ 


_ 


_ 


_ 


_ 


S-C 


CW-NW 


25-50 


2. C. 


seorsiflora 




± 


- 


- 


- 


- 


- 


S 


S-SE 


50-75 


3. C. 


candieans 




- 


- 


— 


- ■ 




+ + + 


S-C 


CW-NW 


25-100 


4. C. 


setosa . . 




+ 


- 


- 


- 


- 


- 


c 


CW 


100 


5. C. 


androstemma . 




+ 


£ 


- 


+ 


+ 


- 


s 


CW 


37 


6. C. 


petrophilovles . 




+ 


± 


- 


- 


- 


- 


s 


C-SE 


25-50 


7. C. 


vaginala 




+ + + 


± 


- 


- 


+ 


- 


s 


S-SE 


45-50 


8. 0. 


setigem 




+ 


- 


+ + + 


- 


- 


- 


S-C 


s-sw-cw 


75-100 


9. C. 


juncea . . 




+ + 


± 


- 


± 


- ± 


-■ 


s 


CW 


75-100 


10. C. 


bealiana 




+ 


+ + 


- 


+ + + 


- 


- 


s 


SE 


50-75 


11. C. 


caricina 




+ + + 


+ 


- 


+ + 


- 


- 


c 


CW 


100-110 


12. C. 


aculeata 




+ + 


+ + + 


+ + 


+ 


- 


- 


S-C 


NW-CW- 


30-125 




















SW-S 





*S = sand, siliceous or calcareous; C = clay, including lateritic soils. 

t Refers to the following broad areas in the South-West Vegetation Province: C = central, NW = north-west, 
CW = central-west, SW = south-west, S = south, SE = south-east. 

of the foliar anatomy separate them from other groups of species. Of the remaining 
species examined the following grow commonly in sandy situations, but have been 
recorded also in clayey soils: C. stylidioides, C. candieans, C. setigera and G. aculeata; 
once again, no anatomical character or group of characters separates these out as a 
group. The same situation exists with regard to geographical distribution and rainfall. 
Anatomical characters are therefore genetically controlled and are not modified to any 
great extent by the environment. 

Systematics. 

In many cases anatomical characters have proved closely correlated with those 
determining systematic aflinity. This is to be expected, since the exomorphic characters 
employed in determining species boundaries frequently reflect some internal anatomical 
feature. The following groups of species, shown as being anatomically related in the 
foregoing descriptions, are recognized by Bentham (1873) as being closely related on 
exomorphic characters: 

1. G. stylidioides, C. prolifera, C racemosa. 

2. G. candieans, G. dealbata. 

3. G. setigera, C. melanopogon, G. psyllium. 

4. G. juncea, G. involucrata. 

5. C. aculeata, C. bracteata. C. bromelioides, G. preissii, G. spinuligera, G. serrulata, 
G. laxiflora, G. cymosa. 



206 THE GEKUS CONOSTYLIS. I. 

Key to the Species examined, based on Anatomical Characters. 
1 . Palisade continuous. 

2. Vascular bundles united in small groups by the bundle sheaths 8. C. setigera. 

2.* Vascular bundles discrete. 

3. Epidermal cell walls thickened and lignified. 
4. Leaves terete. 

5. Tannin cells occurring in the mesophyll. 

6. Epidermis ± smooth ; prominent subepidermal cells in the palisade 

; ■ 9. C. juncea. 

6.* Epidermis undulate ; subepidermal cells absent 5. C androstem.ma. 

5.* Tannin cells absent 7. C. vaginata. 

4.* Leaves flat. 

7. Prominent subepidermal cells in the palisade ; leaves often ± terete . . 9. C juncea. 
7.* Palisade of chlorenchyma cells only. 

8. Margins notably square in section ; subepidermal fibres absent .... 4. C setosa. 
8.* Margins rounded ; prominent rows of subepidermal fibres .... 6. C petrophiloides. 
3.* Epidermal cell walls thin and mostly parenchymatous. 

9. Epidermal cells adjacent to vascular bundles lignifled, otherwise parenchymatous .... 

2. C. seorsiflora. 

9.* Epidermal cells entirely parenchymatous. 

10. Epidermis bearing a heavy grey tomentum of branched trichomes 

3. C. candicans. 

10.* Trichomes sparse or absent on leaf surface 1. C. stylidioides. 

1.* Palisade interrupted by an extension of the fibrous bundle sheaths, commonly of alternate 
bundles. 
11. Alternate pairs of vascular bundles united in a sclerenchymatous girder passing across 

the section from epidermis to epidermis 12. C. aculeata. 

11.* Vascular bundles discrete. 
12. Leaves terete. 

13. Epidermis smooth ; prominent subepidermal cells in the palisade 9. C juncea. 

13.* Epidermis iindulate or grooved. 

14. Epidermis with grooves adjacent to alternate vascular bundles 

5. C. androstemma . 

14.* Epidermis with one or two grooves only, adjacent to fibrous interruptions of 

palisade 7. C. vaginata. 

12.* Leaves flat. 

15. Tannin ceils occurring in the mesophyll. 

16. Epidermal cells ± isodiametric ; palisade of 2 cell layers 10. C. bealiana. 

16.* Epidermal cells tangentially compressed ; epidermis of 3 cell layers 

11. C. caricina. 

15.* Tannin cells absent. 

17. Prominent subepidermal cells in the palisade 9. C. juncea. 

17.* Palisade of chlorenchyma cells only 6. C. petrophiloides. 

Acknoivledgements. 

The present study was commenced in the Botany Department, University of 
Western Australia, as part of a Thesis presented for tlie degree of Master of Science. 
The writer wishes to thank Professor B. J. Grieve and members of his Department for 
helpful advice during the initial stages of the project. 

Special thanks must go to Associate Professor G. L. Davis and Professor N. C. W. 
Beadle of the University of New England for advice and helpful criticism during the 
continuation of the study and in the preparation of the manuscript for publication. 

References. 
Abber, a., 1925. — Monocotyledons. A Morphological Study. (Cambridge Univ. Press.) 
Eentham^ G., 1873. — Flora Australiensis. Vol. 6: 428-41. (Reeve: London.) 
DiELSj A., 1906. — Die Pflanzenwelt von West-Australien Sildlich des Wendekreises. (In 

Engler & Drude : Die vegetation- der Erde. VII.) (Engelmann : Leipzig.) 
EsAu^ K., 1953. — Plant Anatomy. (Wiley: New York.) 
GreeNj J. W., 1958. — Cytotaxononiic studies in the Haemodoraceae. M.Sc. Thesis. (Univ. of 

W. Aust.) 
JoHANSEN, D. A., 1940. — Plant Microtechnique. (McGraw-Hill: New York.) 



207 



SOMATIC HYBRIDIZATION BETWEEN PUCCINIA GRAMINIS VAR. TRITICI AND 

PUCCINIA GRAMINIS VAR. SECALIS. 

By I. A. Watson and N. H. Luig, The University of Sydney. 

[Read 24th June, 1959.] 



Synopsis. 
Somatic hybrids between Puccinia graminis var. iritici and P. yramims var. secalis have 
been obtained under glasshouse conditions. Their pathogenic properties combine those of their 
parental strains. It is suggested that somatic hybridization between the two varieties of 
P. graminis may play a role in the origin of new strains of wheat stem rust in the field. 



It has now been well established that when certain selected cultures of Puccinia 
graminis Pers. var. tritici (Eriks. and E. Henn. ) are mixed in the uredial stage on 
plants of the appropriate graminaceous host, hybridization will readily occur (Watson, 
1957; Watson and Luig, 1958). It has been shown that under Australian field conditions 
certain pathogenic strains of P. graminis var. tritici could have arisen as somatic 
hybrids between previously existing strains (Watson & Luig, 1958). Since P. graminis 
var. secalis is widespread throughout Tasmania and was found from widely separated 
areas of the eastern Wheat Belt of Australia in 1958, the possibility of somatic hybrids 
occurring between this organism and P. graminis var. tritici has been studied. It is 
well known (Stakman et al., 1930) that these two varieties of P. graminis will 
hybridize sexually on the barberry. 

Table 1. 

ReaHion Types of Twelve Wheal Varieties. Rye and Agropyron repens to Parental Strains of P. graminis 

and Two Hybrids between Them. 





Parents. 


Hybrids. 




Red 


Orange 






Variety. 


P. graminis 


P. graminis 








var. secalis 


var. Iritici 


M9-a. 


MlO-a. 




57241. 


5rR2. 






Little Club . . 


;i, 2 - 


4 


3C 


3C+ 


Marquis 


; 


3 + 


;2 = 


x,2+ + 


Reliance 





3 + 








Kota 


0; 


3 


0; 


0; 


Arnautka 


0; 


0; 


0; 


0; 


Kubanka 


0; 


X 


0; 


0; 


Acme 


0; 


0; 


0; 


0; 


Einkorn 


;1- 


3 + 


3-c 


3- 


Vernal 


; 


3 + 


;i + 


3 


Khapli 


0; 


; 


0; 


0; 


Yalta (Sr 11, Sr 12) 


; 


3 + 


3 


;i,2 = 


Morocco W1103 


;1 


3 + 


;1 


3 + 


Black Winter Rye . . 


l+,2,3 + 


;1 = 


1,2,3 + 


1,2,3 + 


Agropyron repens 


2 


0; 


0; 


0; 



Using single spore cultures of red P. graminis var. secalis (accession number 
57241; from Gosford, N.S.W.) and orange NR-2 (Watson, 1957) and adopting the same 
crossing procedures as previously outlined (Watson, 1957), two different somatic 
hybrids were obtained from several mixtures. The reactions of these hybrids on 
varieties of Table 1 which have been stable over several generations are clearly 
distinguished from those of any existing Australian strain of P. graminis. M9-a, for 



Proceedings of the Linnean Society of New South Wales, 1959, Vol. Ixxxiv, Part 2. 



208 SOMATIC HYBRIDIZATION BETWEEN PUCCINIA GEAMINIS VARIETIES. 

example, which attacks Yalta is avirulent on Morocco, usually a very susceptible variety. 
Emmer is susceptible to MlO-a, but Kota resistant, and we have not previously found 
that combination in any one strain. Moreover, both strains, although they do not 
produce a very susceptible reaction on Black Winter Rye, are more virulent on rye 
than are the parental strain NR-2 and other strains of P. graminis var. tritici. 
Agropyron repens, which is susceptible to P. graminis var. secalis, is resistant to both 
strains M9-a and MlO-a and Little Club is not fully susceptible to either. Hence the 
two hybrids may be considered more or less intermediate in, their parasitic properties 
between those of their parental varieties of P. graminis. 

Discussion. 
There is considerable significance in these results from the point of view of breeding 
for stem rust resistance. Since from this cross a hybrid was derived having virulence 
on Vernal Emmer there seems no reason why such intervarietal crosses would not 
produce hybrids capable of rendering ineffective other genes now used as sources of 
rust resistance. In this regard P. graminis var. secalis must be considered as an 
avirulent strain of P. graminis var. tritici. 

lAterature Cited. 
Stakman, E. C, Levine^ M. N., and Cotteb, R. N., 1930. — Orig-in of physiolog-ic forms of 

Puccinia graminis through hybridization and mutation. Scientific Agriculture, 10: 707-720. 
Watson^ I. A., 1957. — Further studies on the production of new races from mixtures of races 

of Puccinia graminis var. tritici on wheat seedlings. Phytopath., 47 : 510-512. 
, and LuiG, N. H., 1958. — Somatic hybridization in Puccinia graminis var. tritici. 

Proc. Linn. See. N.S.W., 83: 190-195. 



209 



DIPTERA OF KATOOMBA. PART 3. 

Stkatiomyiidae and Tachinidae. 

By G. H. Hardy. 

[Read' 24th June, 1959.] 



Synopsis. 
Synonymy is recorded and four species that have not previously been found in New South 
Wales are given under Stratiomyiidae. Characters and synonymy are discussed for some 
variable Australian Tachinidae and include three specific names in genus Actia, four g-eneric 
synonyms under Aprotheca, which include seven specific names placed as synonyms, and six 
names are excluded from the generic conception adopted. Keys are given to aid in identifying 
the species of Odontomyia, Actia and Aprotheca. 



STRATIOMYIIDAE. 

Foreicord. — Previous records from the Blue Mountains of New South Wales are 
Metopotiia geniina Hardy, from Leura, OpModesma flavipennis Macq., Blackheatli, and 
Odontomyia opertanea White, recorded in Hardy 1920, 19.32 and 1938. A record of a 
species of Chiroynyza from Blackheath is mentioned in Hardy 1924, but no more 
specimens have been found. 

Eye coloration. — Taxonomists take advantage of the colour patterns in the eyes of 
Diptera wherever this feature occurs, and use it as an aid towards specific isolation. 
Occasionally the pattern is found to be unstable, but conforming in its variation within 
a sequence of development that reaches a maximum that is the typical form for the 
species. Some recorded colour patterns are given in Hardy 1938 and 1939, and in the 
latter paper, the scheme adopted for describing the patterns is given. More of these 
colour patterns are recorded below. 

References. — Earlier papers give references to species and those are not repeated 
below, only subsequent references being added here. 

BoREOiDEs suBULATUs Hardy. 

Hardy, 1920, Peoc. Linn. Soc. N.S.W., 45: 540; 1924, 49: 364. Kershaw, 1926, Vict. 
Nat., 43: 159. 

Hab. — Leura: 12 April 1953. This species is stated to infest a garden regularly 
every autumn, and some specimens and eggs were transferred to my garden at 
Katoomba, but failed to become established. 

AcTiNA BKEViHiRTA Hardy. 

Hardy, 1932, Proc. Roy. Soc. Qland, 44: 42. 

Hah. — Katoomba; 31 males, 30 females. The earliest date in any year is 22nd 
September 1955, and the last 27th March 1958. Many specimens were on an old compost 
heap that was covered by a dense growth of a weed (Ranunculus) during January 1959. 

Described from Queensland, this record is the first from New South Wales. 

AcTiNA iNcisuRALis Macq. 
Hardy, Proc. Roy. Soc. Tasm., 1920, pp. 40-1; 1932, Proc. Linn. Soc. NS.W., 43: 
53-4. Fuller, 1934, Proc. Linn. Soc. N.S.W., 59: 190-6. 

Hab. — Katoomba; this common species is occasionally seen. 

Neoexaireta spinigera Wiedemann. 
Hardy, Proc. Roy. Soc. Tasm., 1920, pp. 42-3; 1939, Proc. Linn. Soc. N.S.W., 64: 37. 
Malloch, 1928, Proc. Linn. Soc. N.S.W., 53: 361-2. 
Hab. — Katoomba; very abundant. 

Proceedings of the Linnean Society of New South Wales, 1959, Vol. Ixxxiv, Part 2. 

E 



210 DIPTERA OF KATOOMBA. Ill, 

J 

Damaeomyia nitens Hardy. 
Hardy, 1931, Ann. Mag. Nat. Hist. (10), 8: 125. 

Hab. — Katoomba; 1 male, 1 female, 20th and 19tli November 1953 respectively. 
On the male the eyes were reflecting red-green, and the female had in addition a 
red bar at antennal level. 

Damaeomyia tasmanica Kertesz. 
Hardy, 1931, Ann. Mag. Nat. Hist. (10), 8: 125-6. James, 1950, Proc. Ent. Soc. 
Washington, 52: 312-5. 

Hab. — Katoomba; 1 female, 19 November 1957. 

Damaeomyia clivosa Hardy. 

Hardy, 1931, Ann. Mag. Nat. Hist. (10), 8: 128. 

Hab. — Katoomba; 1 male, 8 January 1958; 1 female, 2 February 1959. Known 
previously from Brisbane to Sydney, the present record at over 3000 feet suggests that 
the species will be found widely distributed southwards into Victoria. 

Normally this species has the scutellum raised 30° to the plane of the thorax. 
Contrary to a prior statement (Hardy, 1950) within this genus the scutellum does 
articulate, rising and falling with the wing motion. The tests were made with the flat 
type of scutellum, and now these two specimens were tested for the raised scutellum. 
In both cases tests showed a considerable stiffness existed, especially on this female, 
and it became necessary to depress the scutellum; then the wing action became apparent. 

Acanthasargus palustbis White. 

Hardy, Proc. Roy. Soc. Tasm., 1920, p. 48. 

Hab. — Katoomba; .8 males, 4 females, November to January during the years 1951 
to 1959. Described from Tasmania, this record is the first from the mainland of 
Australia. 

When alive the eyes showed, in accord with age at time of capture, red-green 
reflections that developed red above antennal level and entirely green below, thence 
green above also, leaving a red bar just above antennal level. A silvery sheen of the 
abdomen on the male is seen from the front and side views. 

AcANTHASAEors FLAViPES Hardy. 
Hardy, 1932, Proc. Roy. Soc. Qland, 44: 48-9. 

Hab. — Katoomba; 1 male, 23 December 1955. Described from Queensland, this 
record is the first from New South Wales. 

Ophiodesma flavipalpis Macq. 

Hardy, 1932, Proc. Roy. Soc. Qland, 44: 44. 

Hab. — Katoomba; 4 males, 2 females and several others of inferior condition. 
September to November, 1957 and 1958. 

One male had two red bars below and one above antennal level, and above these 
was a blotch mixed red and green, each reflecting strongly in accord with the direction 
of light. A female had a red blotch both above and below the two central bars, and 
the red area on a third specimen was reduced to two bars both above and below 
antennal level. These differ considerably from eye markings of the allied species 
0. innoda Hardy 1932. 

Genus Odontomyia Meigen. 
So variable are the characters of two common Australian species, that it proved 
impossible to discover features that may distinguish between soine forms given specific 
names. There is a trend in any single breeding area to yield a definite variant which 
may be an ecological form, but long series collected over wide areas are found to have 
characters grading from one described form to another. White (1916) recognized 



BY G. H. HARDY. 211 

these four, but separated carinifacies on colour marks and the style of the antennae 
being attenuated, stating the species appears to be scarce. The attenuated antenna! 
style is an abnormality sometimes found on both sexes in a long series collected from 
a single area. It has been found from Tasmania to New South Wales, and probably 
is commonest in colder regions of these States. 

Key to species of Odontomyia. 

1. Scutellar spines curved to become upwardly directed 2. 

Scutellar spines straight, lying horizontally in continuity with the scutellum 3. 

2. Scutellar spines arising subapically from beneath the scutellum opertanea "White. 

Scutellar spines arising from the apical margin of the scutellum scutellata Macq. 

3. Abdomen with the dorsal area black with green or yellow side spots varying in size ; when 

large they are generally confluent at the lateral margin hunteri Macleay. 

Abdomen with dorsal area black centrally, with a lateral green or yellow border, varying 
in width decipiens Guerin. 

Odontomyia decipiens Guerin. 

Oxycera decipiens Guerin 1838. Hardy, 1938, Peoc. Linn. Soc. N.S.W., 63: 70. 

Synonymy. — amyris Walker 1849, annulipes Macquart 1849, ialmenus Walk. 1849. 
kirchneri Jaennick 1867, marginella Macq. 1849, pectoralis Thomson 1869, picea Walk. 
1850, regisgeorgii Macq. 1838, ruftfacies Macq. 1849, stylata Macq. 1847, subdentata Macq. 
1849. 

0. pallida Hill, 1919, Proc. Linn. Soc. N.S.W., 44: 456, is a new synonym. 

Hab. — Although widely distributed over coastal and lowland areas of Australia, 
this species is unknown from the mountains of New South Wales. 

Hobart; 1 female, 8 January 1955, had its eye colour reflecting red and green with 
a red bar at antennal level, and not quite reaching the posterior border of the eye. 
Above this bar the red dominated, and below it the green had the stronger reflections. 

Odontomyia hunteri Macleay. 

Stratiomys hunteri Macleay 1830. Odontomyia hunteri Hardy 1938, Proc. Linn. 
Soc. N.S.W., 63: 72. 

Synonymy. — Carinifacies Macq. 1849, including var. grandimaculata and var. minima 
Hardy 1920, laterimaculata Macq. 1849, stricta Erichson 1842, and sidneyensis Schiner 
1868. 

0. ohscura Hill, 1919, Proc. Linn. Soc. N.S.W., 44: 457, is a new synonym. 

Hah. — Katoomba; 3 males, 22nd, 27th and 28th December 1955. The last, on one 
antenna only, had the style attenuated as on White's specimen of carinifacies. 2 females 
December t949 and 9th November 1950, also a series of both sexes which are inferior 
in condition. 

Tasmania; Hobart, 7th and 11th January 1955. The male has a red bar which was 
below antennal level near the face, and it sloped upwards. The female has a purple 
bar at antennal level and a large purple blotch above on the otherwise green eye. 
Eight other males and one female were captured in the same month, and all specimens 
are typical laterimaculata of White's interpretation. 

Odontomyia scutellata Macq. 
Hardy, 1938, Proc. Linn. Soc. N.S.W., 63: 72. 
Hab. — Katoomba; 1 male, December 1949. 

References (Stratiomyiidae). 
FULLES, 1934. — Proc. Linn. Soc. N.S.W., 59: 190-6. 
HardTj 1920. — Proc. Roy. Soc. Tasmania, pp. 33-64. 

, 1920.— Proc. Linn. Soc. N.S.W., 45 : 532-542. 

, 1924. — Idem, 49: 360-370. 

, 1931. — Ann. Mag. Nat. Hist. (10), 8: 120-9. 

, 1932. — Proc. Roy. Soc. Queensland, 43 : 50-5 ; 44 : 41-9. 

, 1938. — Proc. Linn. Soc. N.S.W., 63: 70-4. 

, 1939. — Idem, 64: 34-40. 



212 • DIPTERA OF KATOOJilBA. Ill, 

Hardt, 1950. — Ent. mon. Mag., 86 : 230. 

Hill, 1919. — Proc. Linn. Soc. N.S.W., 44: 458-462. 

Jambs, 1950. — Proc. Ent. Soc. Washington, 52 : 312-5. 

Kershaw, 1926. — Vict. Nat., 43: 159-160. 

Malloch, 1928. — Proc. Linn. Soc. N.S.W., 53: 361-5. 

TACHINIDAE. 
Foreword. — The two genera Actia and Aprotheca are parasites of Lepidoptera that 
were becoming of interest to economic entomology, and the endeavour to unravel the 
complicated features in literature concerning them was first undertaken by me when 
holding the Walter and Eliza Hall Fellowship in economic biology at the Queensland 
University (1922-34). The work was continued later, extending into the period of my 
retirement to the Blue Mountains of New South Wales. The experience gained from 
field observation on these flies, under two quite different climatic conditions, has 
permitted a better understanding of the genera than had been acquired before. There 
remains now only the confirmation, or otherwise, of identity by comparison with 
surviving types of the Australian Linnaemyiini in European collections, and that 
were described by very early authors; apparently none were described by them 
belonging to genus Actia. The synonymy given below appears to be an accurate 
interpretation of conclusions reached during recent researches on these genera. 

Tribe Actiinae. 

During recent years, the trend has been to amalgamate various Calyptrate families 
into one family unit, Tachinidae, and to divide the complex formed into divisions that 
do not follow the traditional classification. These and other proposed classifications for 
the family can be regarded as tentative efforts to improve the taxonomy of the Diptera, 
efforts that ultimately may achieve an advancement, but at present the schemes become 
too elaborate in their effect to render the phylogeny of the Calyptratae on a satisfactory 
plan. In these notes Tachinidae is a name retained in the original sense, and in this 
grouping it becomes divisible into three easily identified parts. 

The subfamily Phasiinae has the broadly visible sternites of the abdomen; the 
Dexiinae has the aedeagus of the male divided into two articulating parts, the second 
of which is very long and slender, and finally the Tachininae incorporates the remaining 
genera and is divisible into tribes. All three are separated from other Calyptrates that 
may have the strongly upturned vein M,, by the postscutellum which bulges below the 
scutellum without the concavity between them, a concavity such as seen on Calliphoridae 
and Sarcophagidae. The two following tribes belong to the Tachininae; the 
Linnaemyiini has some genera with minute palpi and, according to Malloch, only one 
genus of Actini occurs in Australia, and is based largely on vestiture. Vestiture is an 
unsatisfactory feature to use because of variability within species and leads either to 
assembling species into complex units or to forming monotypical genera and subgenera 
to accommodate the less usual forms. Already the use of it has led to much synonymy 
needing clarification by the aid of field observation and breeding. 

Genus Actia Desvoidy. 

Between 1929 and 1936, J. R. Malloch had recorded 14 specific names for about 70 
specimens of the genus Actia. The characters used in antennal proportions, chaetotaxy 
and other features, are proving to be so variable on some species that it becomes 
necessary to seek new methods for determining the specific limits. It had been found 
necessary to reduce two of his names to synonymy under A. darwini Mall. (Hardy, 
1938), a species that occurs abundantly around Brisbane, and below three more names 
are similarly reduced under fergusoni Bezzi, which species is abundant in Katoomba. 

The surprising feature in the synonymy is A. eucostnae Bezzi, which proves to be 
an aborted female specimen of A. fergusoni, occurring in about 3% of known specimens. 

The following list gives the number of specimens in parentheses that were studied 
by Malloch. 

A. fergusoni Bezzi 1923, c? (12, both sexes mentioned by Malloch); eucosmae Bezzi 
1926, 5 (2, = fergusoni, aborted form); valida Curran 1927, c?. ? (2); norma Mall. 1929, 



BY G. H. HARDY. 213 

c?, 2 (25); darwini 1929, cJ, $ (5); invalida J", J and sp. J, 1930 (7 + 1, = fergusoni) ; 
haldwini ,$ (1); lata S (1); parviseta ^ (2); ftrems cJCl or ?2,in the Ferguson collection 
one female is marked allotype, but no record is given for this, = darwini) ; iiigritula $ 
(4); argentifrons c? (D; plebia ^, J (7); guadriseta 1936, 5 (1, = darwini). 

Key to species of Actia. 

1. AVith fissicorn third antennal segment on male only 2. 

With normal antennae on both sexes 3. 

2. Upper branch of third antennal segment on male with forked subbranches. 4 posterior 

postsutural dorsocentral bristles baldivini Mall. 

Upper branch of the third antennal segment simple, and second segment of the arista, 
elongate on both sexes. Normally with 4 postsutural dorsocentral bristles, but sometimes 
only three are apparent. Abdomen brown on male and on female normally black with 
segmentations brown and white. They vary towards the first two segments on the female, 
being largely brown, and ventrally the amount of brown varies too. Always the abdomen 
is slightly shining fergusoni Bezzi. 

3. Main radial vein bare, only the radial sector is setulose 4. 

Both branches of the radial field' setulose 6. 

i. With 4 posterior dorsocentral bristles. Abdomen black parviseta Mall. 

With 3 posterior dorsocentral bristles. Abdomen yellowish-brown, varying to quite dark . . 5. 
:i. Anal vein complete norma Mall. 

Anal vein incomplete lata Mall. 

C. With 3 posterior dorsocentral bristles. Largely yellowish species darwini Mall. 

With 4 posterior dorsocentrals 7. 

7. Abdomen- all black on female (male unknown) nigritula Mall. 

Abdomen yellowish at sides anc/ustifrons and plebia Mall. 

AcTiA FEEGUsoNi Bezzi. 

Schizotachina fergusoni 'Bezzi, 1923, Proc. Linn. Soc. N.S.W., 48: 157, fig. 8; 
Tcwnsend. 1926, Philippine J. 8ci., 29: 542 i Schizoceromyia) ; Malloch, 1929, Proc. Linn\ 
Soc. N.S.W., 54: 116, and 1930, ibid., 55: 304, fig. 32A. Actia eucosmae Bezzi, 1926, 
Ann. Mag. Nat. Hist. (9), 17: 239; Malloch, 1929, Proc. Linn. Soc. N.S.W., 54:, 116, and 
1930, iMd., 55: 307. Actia valida Curran, 1927, Ent. Mitt. Berlin, 16: 356 (Schizactina); 
Malloch, 1930, Proc. Linn. Soc. N.S.W., 55: 305. Actia invalida Malloch, 1930, Proc. 
Linn. Soc. N.S.W., 55: 305, fig. 32B. Actia sp. Malloch, 1930, Proc. Linn. Soc. N.S.W., 
55: 305. 

New synonymy. — The description of A. eucosmae agrees with those abortive female 
specimens of fergusoni that have the wing venation incomplete, leaving the upper 
branch of the median field short. Similarly the position of the median cross-vein, 
relative to the radial-median cross-vein, varies in the species and normally is placed 
halfway between r.-m. cross-vein and the bend of the upper median vein, and may 
extend beyond that. 

Chaetotaxy, as used by Malloch, varies too much to be of specific value, and 
similarly species cannot be isolated on the proportions of segments in the arista. 
A. valida Curran and A. invalida Malloch can therefore be regarded as being synonyms 
only. 

When handling some freshly caught specimens in Brisbane, the end of the arista 
on one specimen gradually broke away in pieces, leaving a shorter and shorter remnant 
of the third segment. This suggested that the reduction on specimens may be due to 
wear taking place, but this shortening has not been repeated on test with Katoomba 
specimens. The length of the second segment of the arista varies too, and there is 
quite a complex of relative lengths in these two segments. 

Hah. — Katoomba; numerous specimens, mainly females, collected over the years 
1952-1959, during the months from September 8th to May 17. The last date includes 
an aborted female specimen agreeing with eucosmae Bezzi, and another of these is 
dated 18th April, both in the year 1954. About 3% of known female specimens are 
found to be aborted to more or less the same degree. 

Actia norma Malloch. 
Malloch, 1929, Proc. Linn. Soc. N.S.W., 54: 116; 1930, ihid., 55: 307, fig. 33. 
Hal). — Katoomba; occurring from early June (11th), but rare until September- 
December, and then becomes less plentiful to early February (3rd). 



214 DIPTERA OF KATOOMBA. III. 

Some recorded characters are inconstant, and it is possible that A. lata Mall, may 
prove to be conspeciflc. 

ACTIA PAEVISETA Mall. 

Malloch, 1930, Proc. Linn. Soc. N.S.W., 55: 308; Hardy, 1938, Proc. Roy. Soc. Qland, 
49: 68. 

Hab. — Katoomba; 2 females, the allotype and paratype J, 5 January 1959. The 
species was based on two males, and in accord with key characters, these two female 
specimens can be regarded as only conspeciflc. They are black specimens with very 
little brown, varying in position and amount on legs of the two sides on one 
specimen, but the other has entirely black legs. The white pulverulent covering on 
the abdomen gives a very slight whitish reflection, but forms a dense complete line of 
wliite at the abdominal segmentations. In general appearance the species comes nearest 
to fergusoni, but has a distinctly greater dorsal arch in its outline from head to 
abdominal tip, and it was distinguishable thereby when seen in the fleld. 

Tribe Linnaemyiini. 

Taxonomy. — The species of this discussion differ from other Tachininae by having 
minute palpi. The typical genus Linnaemyia has frontal-orbital bristles on the male, the 
parafacials and eyes hairy, and the third antennal segment is about twice as long as 
the second. In these characters 0. flavipennis Macq. {- similis Walk.) agrees; however, 
the species has been placed as genotype of both Amphibolosia Surcouf and Ballardia 
Curran. 

It is usual to regard this species and M. hrevigaster Macq. as being congeneric, 
the latter name being quoted as genotype of Chaetophthalnius Brauer and Berg., but 
an error was made in assuming that the genotype had frontal-orbital bristles on the 
male brevigaster, misleading some subsequent authors. 

All proposed genera concerned with the synonymy given here are monotypical, 
and when a series of genera are each based upon a single species, all in one zoological 
region, the fact suggests that the treatment given has been very faulty. It is advisable 
to amalgamate the generic conceptions of doubtful validity, and make a new approach 
to the problem of speciflc identities. 

There are two common species belonging to the tribe widely spread in eastern 
Australia, and these are very variable in characters. The following names in sequence 
of publication refer to these two species and the description of one other suggests that 
three species occur: 

M. brevigaster Macq. 1846 (c?) and M. bicolor Macq. 1848 (5) are sexes of one. 
A. rufipes Macq. 1849 (c? & ?) has not been recognized again. 0. flavipennis Macq. 1949 
(2), 0. nucUstylum Macq. 1854 ($), T. similis Walk. 1956 (c?), B. pallipes Curran 1927 
(J') and C. biseriatus Mall. 1930 {^) form the third. 

In addition, in 1929 Malloch identified specimens, giving Macquart's two first names 
to forms that did not agree, placing them under different genera. 

M. brevigaster Macq. c? has insufficient data in its description to determine its 
identity alone, but the figure given for the conspecific M. bicolor 5 shows the dark 
colour of the abdomen very wide and applicable only to forms without frontal-orbitals 
on the male. Macquart's description of M. vittatus ^ (p. 150), which is a Cuphocera, 
is described with frontal-orbitals (deux soies pres du bord internes des yeux), showing 
that Macquart noted this character when present on males, and this followed 
immediately after the description of brevigaster ^ (p. 159) wherein the character is 
not mentioned. Obviously it was absent. 

One main difference gathered from literature lies in the very wide and the narrower 
summit of the head. The very wide case is recorded for flavipennis and nudistylum, 
both are females, and the male of the former is in agreement with the definition of 
Chaetophthalnius as originally but faultily defined. According to Austen, the holotype 



BY G. H. HARDY. 215 

of Tachina similis Walker conforms, as does also a paratype of pallipes Curran. The 
male of diseriatus Mall., in its description, is the same species. 

Habits. — Several hundreds of freshly caught adult specimens have been examined 
over the years 1922-1958, but only three pupae have been discovered by me. These 
pupae were clustered together in the soil without host remains, and from them females 
emerged (5 May 1957). The adults which mainly frequent the ground, low herbage and 
flowering shrubs, get trapped at windows, and they look very like Calliphora subgen. 
Proekon, or, as in one case, subgenus Neopollenia. They may be found every month of 
the year, sometimes in enormous numbers, becoming the dominant Tachinid fly of a 
district for a short time. In the colder localities they are found from early Spring to 
late Autumn. 

Genus Apeotheca Macq. 

Aprotheca Macquart 1849, Dipt. Exot., suppl. 4: 175; Brauer & Bergenstamm 1891, 
Denk. Akad. Wiss. Wien, 58: 408, 444: 1893, ibid., 60: 224. Chaetophthalmus Brauer «• 
Bergenstamm 1891, I.e., 58: 383; 1893, ibid., 60: 145. Amphibolosia Surcouf 1914, 
Nouv. Arch. Mus. Paris, 6: 109. Ballardia Curran 1927, Bull. Ent. Res. London, 18: 166. 
Apalpus Malloch 1929, Peoc. Linn. Soc. N.S.W., 54: 318. 

Synonymy. — Although the genotype of Aprotheca remains unrecognized in Aus- 
tralian collections, its general position in taxonomy remains without doubt. From 
other Australian forms it differs by the absence of the appendix in the radial field of 
the wing. Though usually present, the appendix is found to atrophy and hence 
becomes unreliable as a generic character. 

Under Chaetophthalmus only two names are given in Brauer and Bergenstamm 
(1891-3), namely, Micropalpus brevigaster and M. bieolor, respectively the male and 
female of one species. 

Amphibolosia Surcouf, with genotype Ochromyia flavipennis Macq., is the form 
with an extra wide frons, and frontal-orbital bristles on the male. 

Ballardia Curran has the same genotype under the name pallipes. 

Apalpus Malloch, based on one female specimen, has an oustanding bristle in the 
parafacial region. Where chaetotaxy is found differing so widely even within a species 
this character cannot be regarded as of generic value. The description is based on a 
unique female, yet the figure is labelled male. The sternopleurals are stated to be 
1:1 or 1:1:1, possibly differing on the two sides of the specimen. The inner series 
of frontal bristles is said to extend almost to the eye, but illustrated otherwise on the 
figure. 

Key to species of Aprotheca. 

1. Male with frontal-orbital bristles. Both sexes with summit of head wider than eye-width. 

(Subgen. Amphibolosia. ) similis Walker. 

Male without frontal orbital bristles. Both sexes with summit of head less wid'e than eye- 
width. (Subgen. Aprotheca.) 2. 

2. At least female with an outstanding parafacial bristle. (Male unknown.) Western 

Australian dorsalis Malloch. 

Without such parafacial bristle ;i. 

3. Abdominal stripe normally occupying- much of the tergites dorsally, but not extending far 

along apical margins of the segments brevigaster Malloch. 

Abdominal stripe narrow, but extending into bands along the margins of three tergites .... 
rufipes Macquart. 

Aprotheca rufipes Macq. (1849, p. 176) apparently is not recognizable from descrip- 
tion, and possibly was from Sydney, not Tasmania as recorded. The description comes 
very near to A. brevigaster Macq. and may prove to be a variation of it. 

A. dorsalis Malloch (1929, p. 318) is described as having a narrow frons, and it is 
assumed here that, when found, the frons of the male will be similarly narrow and will 
be without the frontal-orbital bristles. 



216 DIPTEKA OF KATOOMBA. Ill, 

Apeotheca brevigaster Macq. 

Micropalpus brevigaster Macquart 1846, Dipt. Exot., suppl. 1: 149; Schiner 1868, 
Reise Novara, Dipt., 2: 330; Hardy 1938, Proc. Roy. Soc. Qland, 49: 68 (Chaetoph- 
thalmus). Micropalpus bicolor Macquart 1848, Dipt. Exot., suppl. 3: 44; Brauer 189S, 
Sitz. Acad. Wiss. Wien, 107: 495 (.Chaetophthalmus) ; Hardy 1938, Pr'oc. Roy. Soc. Qland, 
49: 68. ? Nemoroea brevigaster Macquart 1849, Dipt. Exot., suppl. 4: 183. ? Linnaemyia 
bicolor Malloch 1929, Peoc. Linn. Soc. N.S.W., 54: 317. 

SynQnym.y. — Malloch attached the first name to a specimen from North Queens- 
land which had frontal-orbital bristles on the male, and this certainly is an error. The 
female from Barrington Tops (N.S.W.) has been discovered at Katoomba and may prove 
to be a variant of brevigaster. The ground-colour of the abdomen sometimes showing 
below the pulverulent overlay suggests this, and in general appearance it looks like 
subgenus Neopollenia (Callipliora) due to that overlay. Moreover the small series 
from Katoomba (1 J', 6 $) shows a graduating density of overlay. Also, some quite 
normal looking specimens of brevigaster have a slight overlay of the same colour, seen 
when viewed from the rear, making the abdomen look brown, and so far no specimens 
showing a denser covering than that have been found. 

Those specimens under Nemoroea added to the original series by Macquart are 
recorded with long palpi and hence cannot be congeneric. 

Had. — Katoomba; abundant from November to January, but in the 1957-8 season 
they were not as plentiful as in previous years. 

Apeotheca similis Walker. 

Ochromyia flavipennis Macquart 1849, Dipt. Exot., suppl. 4: 245; Surcouf 1914, Nouv. 
Arch. Mus. Paris, 6: 110, pi. 5, fig. 5 (Amphibolosia) , name preoccupied. ? Ochromyia 
nudistylum Macquart 1854, Dipt. Exot., suppl. 5: 131; Bigot 1877, Ann- Soc. Ent. France, 
7: 260; Brauer 1899, Sitz. Acad. Wiss. Wien, 108: 517 (Chaetophthalmus); Surcouf 
1914, Nouv. Arch. Mus. Paris. (5) 6: 116; Hardy 1938, Proc. Roy. Soc. Qland, 49: 68 
(Amphibolosia) . Tachina similis Walker 1856, Ins. Saund. Dipt., 2: 266; Austen 1907, 
Ann. Mag. Nat. Hist. (7), 19: 332 {Chaetophthalmus). Ballardia pallipes Curran 1927, 
Bull. Ent. Res. London, 18: 166; Hardy 1938, Proc. Roy. Soc. Qland, 49: 68 (Amphibo- 
losia). Chaetophthalmus biseriatus Malloch 1930, Proc. Linn. Soc. N.S.W., 55: 311. 

Synonymy. — The specimens are so abundant and variable, the characters grading 
from one form to another, and no differences detected in the male terminalia, that only 
one species is possible under the five names. There may be a doubt concerning the 
name nudistylum, originally described from Adelaide and placed as a synonym of 
brevigaster with doubt. 

A more certain identification is Tachina similis Walk, which was based on a male 
with the frontal-orbital bristles. 

The sternites of this species may have two rows of strong bristles, one row each 
on the two apical ones, but the character is a variable one, with bristles decreasing in 
number and often absent on the penultimate sternite, and thus the form biserialis Mall., 
based on a single specimen, becomes, congeneric. 

Hab. — Katoomba. Found nearly every month of the year, chiefly in Spring and 
Summer. 31 August and 1 April are first and last normal dates of occurrence, but in 
addition two females occurred on the wing on the 4th and 9th June 1957; also a male 
on 17 June 1958. 

References (Tachinidae). 
AiTSTBN, E. E., 1907. — Ann. Map. Nat. Hist. (7), 19: 332. 
"Bezzi, M., 1923. — Proc. Linn. Soc. N.S.W., 48: 647-659. 

• , 1926. — Ann. Mag. Nat. Hist. (9), 17: 236-241. 

Bigot, J. M. F., 1877. — Ann. Soc. Ent. France, 7: 243-262. 

Beaubr, F., 1898. — Sitz. Acad. Wiss. Wien, 107: 305-446. 

Brauer, F., and Bergenstamm, J. E. von, 1891. — Denk. Akad. Wiss. Wien, 58 : 305-446. 

, 1893. — Ibid., 60: 89-240. 

* These references are concerned with the genus Actia. 



BY G. H. HARDY. 217 



»CUERAN, C. H., 1927. — Ent. Mitt. Berlin, 16: 345-357. 

, 1927. — Bull. Ent. Res. London, 18: 165-176. 

»HaedTj G. H., 193S. — Proc. Roy. Soc. Queensland, 49 : 53-70. 
MacquarTj p. J. M; 1846. — Bivt. Exot., suppl. 1 : 5-238. 

, 1848. — Ibid., suppl. 3: 1-77. 

, 1849. — Ibid., suppl. 4: 5-336. 

, lS54.^Ibid., suppl. 5: 25-156. 

*Malloch^ J. R., 1929. — Proc. Linn. Soc. N.S."W., 54: 107-116, 283-243. 

* , 1930.— /6id., 55: 303-353. 

» , 1936. — Ibid., 61: 10-26. 

SCHINER, J. R., 186S. — Reise Novara, Dipt. (2): 3-388. 
SuRCOUF, J. M. R., 1914. — Nouv. Arch. Mus. Paris, 6: 27-120. 
*TowN-SHND^ C. H. T., 1926. — Philippine J. Sci., 29 : 529-544. 
"Walker, F., 1856. — Ins. Saund. Dipt., 5 : 415-474. 



218 



SOME EAST AUSTRALIAN SEA-GRASS COMMUNITIES. 

By E. J. Ferguson Wood.* 

(Two Text-figures.) 

[Read 24th June, 1959.] 



Synopsis 

The common sea-grasses in the estuaries of south-east Australia are Posidonia spp 
(probably two species), Ruppia maritima, Halophila ovalis, Zostera capricorni, Z. muelleri, and 
Z. tasmanica. In Z. muelleri there are fewer anthers than in the other two species, which 
may be disting-uished by the spathes. 

In the estuarine environment, Buppia g-rows in water of lower salinity than the other 
grasses, Halophila grows in deep, muddy water, but may be found growing sparsely in the 
shallows, Posidonia normally grows on sandy slopes, but may occur in shallow waters, while 
Zostera grows best on shallow flats with a mud bottom, though the mud may be overlain by 
sand'. Z. muelleri grows near the shore, in shallow water, from Queensland into eastern 
Victoria and possibly further west, Z. capricorni grows from about Noosa in Queensland to 
Lake Tuross in New South Wales, and Z. tasmanica from Tuross south. 

Swans cause great depredations on Zostera beds by digging up and eating the stolons 
and young shoots ; these depredations may be semi-permanent or permanent, depending on the 
water movement over the flats. 



Introduction. 
The taxonomy of Australian sea-grasses has been dealt with by several workers, 
but no ecological studies have been made. This paper deals briefly with the taxonomy 
of the common sea-grasses of south-east Australia and records field observations. 
Numerous attempts have been made to cultivate Zostera and Posidonia under aquarium 
conditions. Even with the simulation of natural conditions, including tidal action, the 
plants have not lived for more than a few months. Sections of the bottom with the 
plants undisturbed v/ere transferred to plastic tanks in running water, but even under 
those conditions the plants grew for only three months, flowered, and died. They did 
not regenerate from the stolons nor did the seeds mature. Owing to these failures, it 
has not been possible to confirm field observations by laboratory experiment. 

Previous Studies. 

Bentham (1877) listed from Australia 7 genera of marine grasses, including the 
following species: Ruppia maritima L., Posidonia australis Hook., Zostera nana Roth., 
Z. tasmanica G. v. Mart., Cymadocea antarctica (Labill.) Endl., C. ciliata (Forsk. ) Ehr., 
C. serrulata Aschers., C. isoetifolia Aschers., Lepilaena australis Drumm., L. cylindro- 
carpa Benth., L. preissi F. Muell., Naias major All., N. tenuifolia R. Brown, Halophila 
ovalis Hook., and H. spinulosa Benth. None of Bentham's records were from New 
South Wales. 

Ostenfeld (1929) gave a more recent list, and some further localities. He listed 
Cym.adocea angustata Ost., C. antarctica (Labill.) Endl., C. ciliata (Forsk.) Ehr., G. 
isoetifolia Aschers., C. rotundatal (Ehr.) Aschers. and Schweinf., C. serrulata (R. Br.) 
Aschers., Diplanthera uninervis (Forsk.), all from Western Australia or Queensland; 
Posidonia australis Hook, from southern Tasmania, and Western Australia north to 
Carnarvon; Zostera capricorni Aschers. from Queensland, Port Jackson, Long Reef, 
Botany Bay, and Harwood Island in New South Wales; and Z. muelleri Irmisch from 
Port Phillip and Point Lonsdale in Victoria, Victor Harbour and Port Pirie in South 
Australia, Tasman Peninsula, D'Entrecasteaux Channel, Port Esperance, and Southport 
in Tasmania. Halophila decipiens Ost. is recorded from Sydney Harbour; H. ovalis 

* Division of Fisheries and Oceanography, C.S.I.R.O., Cronulla, Sydney, Australia. 
Proceedings of the Linnean Society of New South Wales, 1959, Vol. Ixxxiv, Part 2. 



BY E. J. FERGUSON WOOD. 219 

(R. Br.) Hook, from Port Jackson in New South Wales, Pialba to Cape York in 
Queensland, Geraldton, Swan River, and Rottnest Island in Western Australia, Victoria, 
and Tasmania; and H. spinulosa . (R. Br.) Aschers. from Pialba and Moreton Bay in 
Queensland. 

Thalassia hemprichi (Ehr.) Aschers. was collected from Port Dennison and Murray 
Island in Queensland, and Enhalus acroides (L. fil.) Steud. from Thursday Island and 
Cape York. 

Taxonomy. 
' Genus Halophila Thou. 

Submerged marine herbs. Flowers unisexual, solitary, within a pair of herbaceous 
■bracts. Male flower: perianth of 3 segments; 3 anthers, sessile, alternating with 
segments, erect, two-celled, c^ls opening outward; pollen confervoid. Female flowers: 
no perianth, single ovary tapering into a filiform style with a short stigma, either 
single or divided into 3 to 5 filiform segments. Several ovules, erect, attached to sides 
of cavity. Fruits membranous, opening irregularly. Seeds nearly globular, with thin 
testa, rather loose. Embryo erect with thick radicular base nearly the shape of the 
seed, distinct plumula and an involute or spiral cotyledonous end, both nearly immersed 
in a terminal groove. Leaves in pairs, apparently opposite, sessile or petiolate, the 
petioles frequently enclosed at base in 2 broad, scarious, white or hyaline scales. Floral 
Tsracts axillary, sessile, or males pedicillate. 

Bentham states (1877, p. 182) that the genus is confined to the Indian and West 
Pacific Oceans and the Mediterranean Sea. H. ovalis is the common species of the genus 
in the environment considered in this paper. 

Halophila ovalis Hooker. 
Stems creeping and rooting under water, emitting at each node 2 broad, thin, colour- 
less, hyaline scales 2 or 3 lines in diameter. Within these scales lie a pair of leaves 
with long slender petioles and a herbaceous lamina varying from oval and less than 
half an inch long to oblong-elliptical and 2 to 2* in. long, very thin, penniveined, with 
a broad central nerve and very fine lateral, oblique veins; margin quite entire. 
Involucres or double spathes enclosing fiowers ovate, sessile within scales. Male flowers 
on pedicel emerging from involucre, females sessile within it. 

Genus Ruppia L. 

Normally subsaline aquatic plant with slender, much-branched stems, and linear- 
filiform leaves. Flowers hermaphrodite, in a spike enclosed when young in the sheathing 
bases of the floral leaves. Perianth absent. Anthers 2, each with 2 distinct cells 
(described sometimes as 4 one-celled anthers), the cells opening outwards. Pollen 
grains narrow-oblong, slightly curved with ends somewhat dilated. Four carpels, at 
first sessile, but soon protruded on long stalks, each with a pendulous ovule and 
terminating in a short style or broad, almost sessile, stigma. Fruiting carpels ovoid or 
pear-shaped, often oblique, obtuse or more or less produced into a slightly curved beak. 

Bentham (1877), from whom the above description was taken, considers that there 
is only one species which exhibits considerable variation, but not in diagnostic 
characters. The material collected during the present investigation agrees closely with 
his description of R. ma7'itima. 

Ruppia makitima L. 
Stems and leaves submerged, filiform, the leaves often very long; barren leaves 
alternate, slightly dilated at base. Leaves crowded in a tuft, dilated at base into thin, 
almost scarious sheaths, closely imbricate and completely enclosing the young spike, 
which soon emerges on a (short or long) spirally coiled pedicel, bringing it to the 
surface of the water. Flowers 2-6, sessile, at first close together, later often distant. 
Anthers less than half line long. Carpels at the time of flowering are not exceeding 



220 



SOME EAST AUSTRALIAN SEA-GKASS COMMUNITIES, 



anthers, but immediately after flowering anthers fall 
lengthen by as much as half an inch to an inch or more, 
in length. 



away, and stalks of carpels 
Ripe carpels one line or more 



Genus Posidonia Konig. 
Marine submerged plants. Flowers hermaphrodite, or the terminal one of each 
spike male or semi-abortive. Perianth absent. Anthers 3 to 6, consisting of a broad, 
almost fleshy connective, with 2 short dorsal cells at base, separated by the broad, thick 
centre. Ovary of one sessile carpel, with 1 erect ovule, terminating in 2, 3, or more 
short stigmatic lobes. Fruit indehiscent, the pericarp thick and fleshy or succulent. 
Seed adnate to one side of cavity; embryo erect, straight, with a rather large plumule 
in a terminal groove. The base of the stem is covered with filamentous remains of old 




Pig. 1. — Floral differences in Posidonia: a, Merimbula form — showing arrangement of 
anthers, and bj showing fertile shoot. c. Port Hacking form — showing arrangement of 
anthers, and d, showing fertile shoot. Drawn by L. Crosby. 

leaf-sheaths. Leaves often very long, breaking off transversely from the persistent 
sheathing base. Scape leafless below inflorescence, bearing at end several spikes, each 
in axis of a short, floral leaf, the peduncle enclosed at the base of 2 sheathing bracts, 
each flower subtended by a small bract and two bracteoles. 

The genus is at present being revised by Dr. Melville at Kew. but there appear to 
be two forms on the east Australian coast, one collected from Merimbula and having 
6 anfhers per pistil and a long slender leaf (Fig. 1, a and 6), the other from Port 
Hacking, having 3 anthers and a shorter thicker leaf (Fig. 1, c and d). The arrange- 
ment of the spadices is also somewhat different as the figures show. 

Genus Zosteba L. 
Marine submerged plants. Flowers unisexual, males and females in alternate rows 
on the membranous rachis of the spike enclosed in the sheathing base of the floral leaf. 
Perianth absent. Male flowers of a single, ovate, laterally attached, one-celled anther; 



BY E. J. FERGUSON WOOD. 



221 



pollen confervoid. Female flowers of a single carpel, gourd-shaped, laterally attached 
near apex, and produced above attachment into a filiform two-branched style. Fruit an 
indehiscent nutlet. Embryo with deep, longitudinal groove forming 2 valves which fold 
over the long, curved, linear cotyledonous end. Creeping rhizome, emitting short 
stems with long, narrow, grass-like leaves separated from their narrow sheathing 
bases by a transverse line. Peduncles axillary or terminal, bearing a single spike 
completely enclosed in the more or less dilated but continuous sheathing base of the 
floral leaf or spathe, which otherwise resembles the stem-leaves. Rachis of spike broad 
and thin, with margins folded inward and bearing flowers and fruits only on inner 
surface. 




Fig. 2. — Differences of Zostera species : a. Anthers of Z. capricorni (Z. tasmanica is 

similar); b. Anthers of Z. muelleri; c. Fertile shoot of Z. cwpricorni; a. Fertile shoot of Z. 

muelleri ; e. Sterile shoot of Z. capricorni ; /, Sterile shoot of Z. muelleri ; g, Fertile shoot of 
Z. tasmanica. Drawn by L. Crosby. 



Setchell (1933) divided Zostera into 3 sections, Alega, which has not been recorded 
from the Southern Hemisphere, Zosterella and Heterozostera, both of which have 
southern representatives. He points out that there is considerable confusion with 
regard to the type material from Australia and New Zealand. 

There appear to be three species of Zostera represented in east Australian waters 
corresponding to Z. capricorni Ascherson, Z. muelleri Irmisch (both in the Zosterella 
section) and Z. tasmanica (Heterozostera) . 

Setchell divides these species on the leaf venation, shape of leaf apex and the 
retinacula (bracts ? arising within the elevated margin of the spathe), though the 
Heterozostera section has other characteristics. However, the examination of several 
hundred specimens shows considerable variation in the shape of the leaf apex and in 
the leaf venation, thus reducing the usefulness of these criteria for diagnosis of the 
species concerned. This variation can be discerned in leaves from one bed, and even 
from individual plants. Examination of material collected from various parts of Port 
Hacking, Lake Macquarie and Lake Illawarra showed that in Zostera capricorni, 20 
per cent, of the uninjured leaf tips were evenly rounded, 23 per cent, regularly truncate. 



222 ■ SOME EAST AUSTRALIAN SEA-GRASS COMMUJJITIES, 

15 per cent, irregularly truncate, and 42 per cent, regularly or irregularly notched. In 
Z. muelleri, 15 per cent, were evenly rounded, 27 per cent, regularly and 13 per cent, 
irregularly, truncate, while 45 per cent, were notched. The denticulation of the leaf 
tips also varied considerably; this may be due to injury rather than to specilic 
differences, especially as the majority of leaves examined had obviously been damaged, 
the whole tip being missing. 

However, a separation of the species can be made using the number and arrange- 
ment of the anthers. In Z. tasmanica and Z. capricorni, the anthers are arranged in 
two closely packed lows (Fig. 2, a), while in Z. muelleri they are alternate and some- 
what spaced (Fig. 2, b). Z. tasmanica has 20-24 anthers per inflorescence, Z. capricorni 
13-20, and Z. muelleri 10-11. The length of the fertile shoot is not very useful lor 
separating Z. muelleri from Z. capricorni, as both are long relative to the plant (Fig. 2, 
c and d). Z. muelleri can be distinguished in the field from Z. capricorni by the fact 
that the leaves of the former are smaller and narrower. A sample from Lake Macquarie 
of 300 leaves of each species gave mean dimensions of: Z. muelleri, 30 x 2 mm., and 
Z. capricorni, 55 x 4-5 mm., though many samples of the latter from other areas would 
have a far greater length, and leaves up to 1 metre long have been observed at Stewart's 
Point, Crookhaven and other places. Zosterella and Heterozostera are separated by 
the situation of the stem bundles and fibre strands, and by the stem and spathes. In 
Zosterella the rhizomes have one lateral bundle on each flank and fibre strands in the 
inner portion of the cortical layer, the stems are horizontal, with very short vertical 
branches, and the spathes are very slightly swollen but conspicuous. In Heterozostera, 
i.e., in Z. tasmanica, there are multiple vascular bundles in the internodes, the fibre 
strands are just beneath the epidermis, the vertical stems are often long (up to 30 cm.), 
and the spathes markedly swollen, and sessile, being almost or quite shielded by the 
sheath of the leaf below (Fig. 2, g). The floral branches of the Heterozostera are very 
hard to find in the field (cf. Fig. 2, / and g), whereas the long, well differentiated, fertile 
branches of the Zosterella section are easy to see (Fig. 2, c, d, e). 

The differences between the three Australian species are summarized thus: 

Z. muelleri. Z. capricorni. Z. tasmanica. 

Anthers 10-11, alternate. Anthers 13-20, in 2 packed Anthers 20-24, in 2 packed 

rows. rows. 

Leaves up to c. 30 x 2 mm. Leaves from 45-100 x 4 5 Leaves from 45-100 x 5 mm. 

mm. 
One lateral bundle on each One lateral bundle on each Multiple bundles in inter- 
side of internode. side of internode. nodes. 
Fibres in inner cortex. Fibres in inner cortex. Fibres near epidermis. 
Spathe obvious, not swollen. Spathe obvious, not swollen, Spathe inconspicuous, swol- 
len. 
Fertile stem longer than Fertile stem longer than Fertile stem similar to 
barren stem. barren stem. barren stem. 

Ecological Relationships. 

Because the east coast of Australia is characterized by a large number of shallow 
coastal lagoons and meandering river estuaries, the sea-grass associations become 
ecologically and economically important. Within the estuaries they form large areas 
which are associated with inshore fisheries and oyster culture. They occur in micro- 
biologically active areas, the roots being in reducing zones while the leaves are in 
oxidized water which, under active photosynthesis, is oxygen-saturated. 

The grasses found on the east coast are, in rising order of importance, Halophila, 
Ruppia, Posidonia and Zostera. Cymado'cea, which occurs in large beds, and washes 
ashore to form banks up to 6 feet deep and many feet wide in the Bunbury-Busselton 
area of Western Australia, is unimportant on the east coast. 

(a) Halophila. 
Halophila is relatively unimportant in the estuaries of eastern Australia, except in 
some relatively deep, still, muddy areas, often below 20 ft., such as Bundabah (Port 



BY E. J. FERGUSOA WOOD. 223 

Stephens), Merimbula Lake, and deeper parts of Lake Macquarie. In these places it is 
usually dominant or associated with red algae such as Gracilaria sp. Where it occurs 
in shallow water it is sparse to subdominant, e.g.. Little Swan Bay (Port Stephens) 
and Gunnamatta flats (Port Hacking). It is then associated with Zostera or, less 
often, Posidonia, together with numerous red algae such as Gracilaria, Laurencia, 
Oigartina, Polysiphonia and Ceramium, and the brown Ectocarpus. Halophila appears 
to be associated with reduced, sulphide-containing muds. The Lake Macquarie muds 
where Halophila was found had Eh values of -50 to -100 mV, and the same limits were 
observed in the Halophila regions in Port Hacking. While Halophila is usually found 
in areas of lower chlorinity, from 18 to 19-0%c, it can also occur, as it does in Port 
Hacking, in water of normal chlorinity, from 19-5 to 20%^. Illumination in the deeper 
regions of growth was of the order of 5 to 10 per cent, of surface values. 

(b) Ruppia. 

Ruppia, often called "swan grass", is usually associated in the east coast estuaries, 
with water of lower chlorinity, from about lG%c as in Swan Bay (Port Stephens) to 
almost potable water, as at Killarney (George's Basin) and upper Lake Conjola. In 
these cases it is dominant. In other cases, e.g., Galgabba Flats in Lake- Macquarie and 
at the western end of Merimbula Lake, Ruppia is associated with Zostera where the 
chlorinity is 17-5-19-5%t,. On Galgabba flats and Lake Erraring, in Lake Macquarie, 
Zostera was dominant until a large flock of swans invaded the area and removed most 
of the Zostera, leaving the Ruppia. (In both Merimbula and Lake Macquarie, heavy 
freshes may cause stratification and a temporary lowering of the chlorinity to about 5%o 
which may favour the growth of Ruppia.) The Eh of the muds on which Ruppia grows 
is usually negative, though at Killarney the surface was oxidized, but a very strong 
sulphate reduction was going on a centimetre below. Ruppia normally has few epiphytes, 
although at times it does become covered with a mat of Ectocarpus, or filamentous 
blue-green algae, diatoms (chiefly Syneclra ulna) being relatively sparse. Ruppia flowers 
in the summer, from November through to February, but seedlings have been observed 
only at a lagoon near Shell Harbour, New South Wales, and at Paynesville in Gippsland, 
Victoria. Ruppia has not been found in water much more than 6 feet deep at mean low 
water, and appears to require good illumination, and little current. 

(c) Posidonia. 

More than one species of Posidonia has been recognized in Australian waters, out 
the taxonomy of the genus is still doubtful. Posidonia has its greatest development in 
the South Australian gulfs and on the west coast of that State. There, the growth is so 
abundant that numerous attempts have been made to use the fibre commercially 
(Winterbottom, 1917). On the east coast of Australia, however, Posidonia is of interest 
merely as forming a community in the sea-grass association, where it normally occurs 
in waters from about 6 feet below mean low water to between 20 and 30 feet. Here it 
is characteristic of sloping, sandy shores, where the bottom has a positive Eh extending 
for several centimetres in depth, the water is clear, and turbidity slight, e.g., in Lake 
Macquarie, Cabbage Tree (Port Hacking), Weeney and Quibray Bays, and on the 
northern slope of Towra Point in Botany Bay. At Merimbula, the Posidonia occurs as 
the dominant form on shallow tidal flats between the causeway and the entrance to 
Merimbula Lake where the Eh is about -f250mV. At times it is found associated with 
Zostera. e.g., on the Belmont flats (La]je Macquarie), and with Zostera and Ruppia, 
e.g., in the shallows of Merimbula Lake. In such cases it does not appear very healthy. 
Posidonia has not been found in exposed waters on the east coast. 

Posidonia usually has a number of epiphytes on the leaves, e.g., diatoms, red algae 
(Melabesia, Ceramium, Polysiphonia, Laurencia), brown algae (Ectocarpus, Col- 
pomenia), the green algae (Vaucheria, Enteromorpha and Chaetoniorpha) , and blue- 
green algae (Oscillatoria, Lyngbya). In addition, the ciliate, Vorticella, the tube-worm, 
Spirorbis, and encrusting bryozoa are numerous. The epiphytic diatoms are less 



224 SOME EAST AUSTRALIAN SEA-GRASS COMMUNITIES, 

important than on the adjacent Zostera but, in general, the same species occur and 
consist of stipitate forms including LicmorpTiora, Navicula, Skeletonema, Plagiogramma, 
Melosira, Grammatophora, Cocconeis and Amphipleura. 

Though Posidonia is normally the dominant form, the community usually contains 
a number of algae, including Sargassum, Gigartina, Gracilaria, and a very large number 
of species of red algae, Dictyota, often CystophyUum (at Wallis Lake), Codium tomen- 
to.ium and C. lucasi, various Caulerpas. 

The leaves of Posidonia contain a nitrogen-containing reducing substance, volatile 
in steam, which reduces cadmium or lead acetate to the metal, but differs in odour 
from that contained in Zostera (Wood, 1953). 

(d) Zostera. 

Of the sea-grass communities, the Zostera community is most important both 
ecologically and commercially. In Europe, Japan, America and Australia, this plant 
is associated with the growth of Crustacea and other larvae, oysters, and other molluscs, 
and is of general occurrence in estuaries in temperate waters. The "wasting disease" 
of this plant in Europe and America, and the consequent failure of various fisheries in 
these countries, showed how close the relation was between the plant and the animal 
life of the estuary (Renn, 1936). 

Zostera grows in water which is marine in character, and has a chlorinity of the 
order of lS,-l%'/<r, though it may, at times, encroach into less saline waters. It grows 
from low tide mark to a depth of about 20 feet and appears to require good illumination, 
as it does not do well in turbid water. The depth to which it grows seems to depend 
on turbidity. The grass can persist in the presence of currents of several knots, as in 
the channel entrances to Lake Macquarie and Sussex Inlet. Zostera has been found 
growing on bottoms with Eh values of +180 (Lake Macquarie, in gravel) to -150 mV 
(Maianbar, in sand and reduced with abundant H^S (0-2n)), and where the pH varies 
between 5-5 in the mud to 9-3 in the water on sunny days near midday. 

The plants have been found with their root hairs in actual contact with particles 
of hydrotroilite (FeSH(OH)). Zostera normally has its root system in a reducing 
environment and produces organic reducing substances (Wood, 1953); it seems probable 
tliat its metabolism is adapted to such environments. The reducing substances occur 
in the epidermis of the leaves and stolon, and in the leaf gland cells. They are 
associated with a pigment which is possibly a digalloyl. 

The Zostera community is rich with plant and animal life. The microbial groups 
in the mud consist of bacteria of the sulphur cycle, including Desulphovibrio, Thio- 
hacillus sp., Th. denitriflcans, Chromatium, Chlorohium, large numbers of heterotrophs, 
including anaerobes, a few naviculoid diatoms, flagellates, and ciliates, including 
anaerobic Colpidiuvi, Euplotes, and rhizopods. The epiphytes are very numerous and 
include a very large assemblage of diatoms, Ectocarpus, Ceramium, Polysiphonia, 
Oscillatoria, Lynghya. and other blue-green algae, while the protozoan Vorticella, and 
the serpulid SinrorMs are common, e.g., in the channel at Lake Macquarie. The 
epiphytes serve as food for mullets and other phytophagous fish, the gastropod Pyrazus 
etc., and the Zostera itself is consumed, and apparently digested, by the nudibranch 
Aplysia, and the garfish {Reporhampus ardelio) . 

Zostera flowers and fruits in the summer, from November to the end of March, and 
flowering occurs in patches, at times small and localized, and at times large and wide- 
spread. Although a careful watch was kept for several years, no seeds or juvenile 
plants were observed except at Paynesville id Victoria, hence it would appear that 
propagation is very largely vegetative. Zostera muelleri occurs mainly in shallow 
water (low water line to about 5 feet) and Z. capricorni in deeper water (from 2 feet 
below low water to below 20 feet). However, the two species may be interspersed. 
Soon after flowering, most or all of the aerial shoots die off, and only the stolons, which 
are subterranean, and a few residual shoots or stools remain. Thus, in autumn, a 
Zostera flat may seem completely bare of weed. The dead leaves either wash out into 
deeper water or ashore, where they form high banks which rot slowly. Cellulose 



BY E. J. FERGUSON WOOD. 225 

digestion of tlie cell walls is very slow, and no cellulose digesting bacteria were isolated 
from Lake Macquarie. However, Imai et al. (1951) have shown that the decaying 
Zostera serves to harbour and nourish a number of microorganisms including hetero- 
trophic bacteria and colourless flagellates. These authors found that, in Japanese 
waters, one of these flagellates, Monas sp., serves as the food of the larval oyster (Ostreo 
gigas), and that the spawning of this oyster coincides with the death of the Zostera 
in Matsushima Bay and Mangoku Inlet. The spawning of the Australian oyster has not 
been definitely correlated with the death of Zostera, but occurs between December and 
March when the Zostera is declining, so the same relation may hold in Australian 
estuaries. Living Zostera harbours a number of microplankton and phytoplankton as 
well as zooplankton organisms and small flsh. Diatoms such as Coscinodiscus, 
Biddulphia, and dinoflagellates such as Exuviaella, Gymnodiniuvi, Prorocentrum, art- 
common in the community, and so are various crustaceans, e.g., Copepods, Mysids, 
Caprellids, which, no doubt, attract the fish into the community, where they also find 
shelter from their enemies. 

Swans may cause great depredations in the Zostera beds. In the cases observed 
large areas were denuded, and this may be assumed to have had repercussions on the 
fishing potential of the areas. Swan faeces were examined and found to consist entirely 
of Zostera, mainly stolons, the leaves being torn off and allowed to float away. Another 
effect of this denudation by swans is that the bottom is no longer held in place by the 
Zostera stolons, and may be moved about by currents and winds, thus becoming oxidized 
and so unfit for future growth of the plant. This has happened over a large portion of 
the Goodwin Sands, Mallacoota, on the "proclaimed" areas of Lake Illawarra and in 
Lake Erraring, an offshoot of Lake Macquarie. 

The larger organisms associated with Zostera include a number of algae, e.g., 
Enteromorplia prolifera, Chaetomorplia, Microdictyon, Golpomenia, Gystophylluvi, 
Godium, Gaulerpa, Dictyota. etc., and some of these during spring or autumn may form 
a felt which covers the Zostera, and may produce anaerobiosis below it. With these 
felts are Ectocarpus, various filamentous red algae, and diatoms, as well as the micro- 
plankton. Such felts at last float off and may cause a nuisance on the estuary or on 
the beaches, and by fouling nets and lines, making it impossible to fish. 

In the Gippsland Lakes, the Zostera beds are said to have been denuded by crabs 
which invaded the lakes after the opening of the entrance, but it is difficult to know 
at this stage whether the crabs, or changes in lake level and salinity due to deforesta- 
tion, were the real cause of this denudation. Zostera is the most abundant sea-grass in 
south-eastern Australian estuaries. At Stewart's Point, it covei's the old bed of the 
Macleay River for about 15 miles in water from to 6 feet deep. Here, the old mouth 
of the river just south of Grassy Head has been closed for many years, and there is 
little or no water movement. The sea-grass persists in water which, in the summer, 
becomes too hot to walk in, and which kills Gracilaria and other algae. There is a 
similar occurrence of Zostera in the Grookhaven River, near Greenwell Point. In the 
channels of Lake Macquarie, Queen's Lake (Camden Haven), Maianbar, and Sus^^ex 
Inlet, Zostera also covers most of the bottom, but in these cases there is a strong 
current of several knots. Sussex Inlet flows into George's Basin and, in the vicinity of 
the drop-over, as in Lake Macquarie, the Zostera is replaced by Posidonia, almost the 
reverse of the situation at Merimbula. In Port Stephens and Botany Bay, the dominant 
sea-grass on the southern shore is Posidonia, which extends up the estuary to Salamander 
Bay in the former and to Towra Point in the latter. Zostera is dominant on the northern 
shores of these estuaries. In Port Hacking, Zostera is dominant on both shores, except 
on the side of the channel near the Marine Laboratory. During the last 20 years, 
especially in more recent years, Po'Sidonia has extended from the northern slope of the 
channel in Port Hacking, and has grown onto the flats, replacing Zostera. This may 
be due to the effect of man, in altering the environment by di-edging, digging the flat 
tor worms, etc., and suggests that Posidonia is less responsive to changes in environment 
than Zostera. 



226 SOME EAST AUSTRALIAN SEA-GRASS COMMUNITIES. 

Zostera has also been observed in the unprotected Bate Bay and otf Tathra, and 
may occur in 20 to 30 feet on sandy-mud bottom in other places. Estuarine Zostera 
is bright green when young, but later turns purple-brown owing to the development of 
accessory pigments while the stolons become woody. The oceanic Zostera, on the other 
hand, remains green throughout the summer, and the stolons remain succulent. 

Zostera capricorni and Z. muelleri were found from Moreton Bay to Tuross Lake 
(Turlinjah), and Z. tasmanica from Tuross southwards. Z. tasviaiiica and, to a lesser 
extent, Z. muelleri were found at Merimbula, Mallacoota, and in the Gippsland Lakes. 

The areas used for culture of the Australian oyster, Crassostrea commercialis, are 
closely associated with the Zostera beds, the distribution being practically that of 
Z. capricorni. though some areas are suited for spatting and others for fattening, and 
the Zostera community does not give a clue in this matter. There is no doubt, however, 
that the biological activity, especially that of the sulphur cycle, which occurs in the 
Zostera community plays an important part in supplying the microorganisms which 
the oyster requires as food and in the release of phosphate, nitrogen compounds, etc. 

Acknoioledgements. 
Thanks are due to officers of this Division, including Prof. L. G. M. Baas Becking, 
for some of the data used in this paper and for helpful discussions, to Mr. L. H. Crosby 
for collecting material and drawing the diagrams, and Messrs. Rochford and Spencer 
for hydrological data. 

References. 
Bentham, a. J., 1877. — Flora Australiensis. Naiadae, Tribe 2, Zosteraceae : 164-182. 
iMAi^ T., Hatanaka^ M., Sato, R., and Sakai, R., 1951. — Ecology of Mangoku Inlet with special 

reference to oyster seed production. Sci. Repts. Res. Inst. Tohokxi Univ., Dl-2 : 137-15G. 
OsTBNFELD, C. H., 1915. — On the geographical distribvition of the sea-grasses. Proc. Roy. Soc. 

Vict, 27: 179-190. 
, 1916. — Contributions to West Australian botany. I. The sea grasses of "V^'^esl 

Australia. Dansk. Bat. Ark., 2 : 6. 

, 1929. — A list of Australian sea-grasses. Proc. Roy. Soc. Vict., 42 : 1-4. 

Rbnn, C. E., 1936. — The wasting disease of Zostera marina. I. A phytological investigation of 

the diseased plant. Biol Bull., 70: 148-158. 
Setchbll, ~W. A., 1933. — A preliminary survey of the species of Zostera. Proc. Nat. Acad. Sci. 

Wash., 19 : 810-817. 
WiNTBRBOTTOM^ D. C, 1917. — Marine fibre. Sth. Aiist. Dept. Chem., Bull. 4. 
Wood, E. .1. P., 1953. — Reducing substances in Zostera. Nature, 172: 91G. 



22T 



THE EFFECTS OF INORGANIC SALTS ON DIVIDING CELLS. 
By Mary M. Hii^divearsh, School of Biological Sciences, University of New South Wales.* 

(Plate viii; two Text-figures.) 

[Read' 24th June, 19.'j9.] 



Synopsis. 
Onion root tips treated with certain inorganic salts showed abnormal cell division and a 
reduction in the number of dividing- cells with time. The addition of a low concentration of 
other salts to the medium neutralized the effect of the single inorganiJc salt. This inhibition 
of the cell division process seems to be due to ion unbalance in the external solution, rather 
than the specific effect of any inorganic ion. 



Introduction. 
A wide variety of organic substances is known to produce abnormalities in mitosis 
in both plant and animal cells, and the effects of some inorganic salts on the cell 
division process have also been described. In 1945 Levan, who treated onion root tips 
with a number of inorganic salts, mostly nitrates, found that solutions of the salts of 
27 different metals produced cytological changes in the cells of the meristem. Galinsky 
(1949) reported the inhibition of cell division in the same material, by solutions of 
NatHPOi, NaHPO,, and KHoPO^. He attributed the observed abnormalities to the 
presence of the phosphate and compared them to the effects produced by colchicine 
and acenaphthene. Using chick cells in tissue culture, Hughes (1952) found that 
abnormal tonicity upset cell division by suppressing spindle formation and produciug 
blocked metaphase. This raises the possibility that cytological abnormalities in the 
presence of inorganic salts may be due to an ion unbalance in the solution, rather than 
to a specific effect of any inorganic ion. When a phosphate buffer was used to control 
the pH in some experiments with onion seedling roots (Hindmarsh, 1953), it was 
noticed that abnormal mitoses were produced in the root meristems, and experiments 
were designed to find whether this was caused by the phosphate or by the univalent 
ions in the solution. 

Materials and Methods. 

Onion seeds were surface sterilized in a saturated solution of bleaching powder for 
5 to 10 minutes, set out in sterilized petri dishes on filter paper moistened with culture 
solution and incubated for 7 days at 22°C. The culture solution was KNO3 00025M, 
Ca(NOJ„ 0-0025M, MgSO, 0-OOlM, and KH2P04 0-0005M (Hoagland and Broyer, 1936 > 
at a pH of 5-3. 

Preliminary examinations of the relationship between root length and the number 
of dividing cells in the root tips were made to find the least variable material for 
cytological investigation. Counts of dividing cells showed that seedlings with roots 
from 6-0 to 11-0 mm. after 6 to 7 days proved less variable than shorter or longer 
roots. The over-all mean for four experiments was 290-5 for 74 roots, and the pooled 
standard deviation within these four groups (calculated from the mean of each group) 
was 67-7. 

Roots were treated by floating seedlings on 20 c.c. of the test solution in 150 cc. 
Erlenmeyer flasks. All the solutions were autoclaved without any appreciable change 
in pH. The few seedlings which sank during the course of an experiment wore 
discarded. 

* This work wa.s done during the tenure of a Linnean Macleay Fellowship at the Botany 
School, University of Sydney. 

Proceedings of the Linnean Society of New South Wales, 1959, Vol. Ixxxiv, Part 2. 



228- 



EFFECTS OF INOKCiANIC SALTS OX DIVIDING CELLS. 



Root tips were fixed in acetic alcohol and stained in 1-0% aceto-orcein. To count 
the number of dividing cells in each stage of division, roots were heated in acidified 
aceto-orcein stain and the meristematic region stained more deeply than the remainder 
of the root. This deeply stained tip was removed just behind the meristeiiiatic zone, and 
cut into 4 pieces in fresh 10% stain. These pieces were squashed into a single cell 
layer by pressure on the coverslip and division stages scored during a systematic 
examination of the whole slide. 

The buffer solution was a mixture of 2 parts Na„HP04.12H„0 (23-9g/l), and 8 parts 
KHoPO, (907g/r) with a final concentration of 0-06M and a pH of 6-2. Roots were 
treated with 006M KCl, 0-06M, 0-03M, and 0-015M phosphate, and with each solution 
combined with HoagJand's culture solution. Controls were grown in the culture solution. 




20 30 • 40 50 

TIME IN HOURS. 



00 6M. , 
70 



60 
FIG. 



Results. 

a. The effect of the buffer solution. There was no evidence of any chromosome 
fragmentation with this treatment, but other cytological abnormalities were produced. 
The number of dividing cells in root tips was reduced by all concentrations indicating 



T.^BLE 1. 



Time 


Phosphate Concentration. 


Hours. 


0-015M. 


0-03M. 


0-06M. 


15 


Jfumerous normal dividing cells. 


Fewer divisions than normal. 
Chromosomes short and thick, 
otherwise normal. 


Blocked metaphase. Few ana- 
phase — all irregular. 


24 


Mainly normal. Few blocked 
metaphase. 


Few divisions, some normal, 
some blocked. 


Ver.v few divisions, all abnormal. 
No anaphase and telophase. 


40 


Mainly Bormal. Some cells 
have short, thick chromo- 
somes. Few blocked meta- 
phase. 


Most cells resting nuclei, but 
some roots have few prophase 
and blocked metaphase. 


All roots have only interphase 
nuclei. 


40+30 
in water 


All normal dividing cells. 


All normal dividins cells. 


All interphase nuclei. 



BY MARY M. HINDMARSII. 



229 



that the onset of the cell division process was prevented (Text-fig. 1). The spindle 
mechanism was also affected and cytological abnormalities varied with the concentration 
of the phosphate buffer and time (Table 1). ij / i: ■■■■' 

The first observed effects were shortening, thickening of chromosomes in otherwise 
normal metaphase and anaphase stages, followed by irregular metaphase and anaphase 
stages with some chromosomes left off the metaphase plate oi' lagging at anaphase. 
Finally all early stages were affected as they were unable to pass metaphase; blocked 
metaphase occurred instead of normal metaphase plate formation (PI. viii, A and B). 
Some binucleate cells formed, but these were rare, suggesting that there is little effect 
on cells which were in anaphase and telophase at the time of treatment. 



CONTROL. 




TIME IN HOURS 



The cells of roots returned to water or culture solution after treatment with 0-06M 
phosphate buffer for 40 hours did not begin division even after 30 hours. Most cells 
appeared to be quite healthy, but some were pycnotic. The roots remained white and 
turgid and were apparently still alive. There was no "stickiness" of chromosomes such 
as was observed during treatment with nitrophenols (Levan and Tjio, 1948; Hindmarsh, 
1951). After shorter treatment at 0-06M and after treatment with lower concentrations 
the roots recovered in water or mineral solution and commenced normal division. 

In roots treated with 0-06M phosphate buffer and mineral solution together, no 
cytological abnormalities were observed. All stages of normal mitosis were found in 
these roots, but there was a gradual reduction in the number of dividing cells with 
time (PL viii, F). This reduction was not as great as with phosphate buffer alone 
(Text-fig. 2). 

b. The effect of a nincjle salt solution. The entry of cells into cell division was 
affected more by 0-06M KCl than by the same concentration of phosphate buffer 
(Text-fig. 2). KCl solution also produced extra contraction of the chromosomes in 
metaphase and anaphase stages, and a few blocked metaphase stages at 15 hours 
(PI. viii, C, D and E). The addition of a culture solution to the KCl solution reduced 
the effect on the number of dividing cells (Text-fig. 2), and in roots so treated 
the chromosomes as well as the division process were perfectly normal (PI. viii, G). 

Discussion. 
The results of these experiments show that the ionic composition of external 
solution can influence the course of the cell division process. 



230 ■ KFFKCTS OF IMOISGAKIC SALTS OX DIVIDING CELLS. 

Chemical inhibition of the mitotic cycle can occur in one of three ways. Some 
chemicals inhibit or destroy the spindle mechanism, leading to the formation of blocked 
inetaphase, abnormal anaphase and telophase in the early stages of treatment, and 
ultimately to the complete absence of anaphase and telophase, and to polyploidy, e.g., 
colchicine. A second group has a similar effect, but also prevents the onset of prophase 
leading to a reduction in the number of dividing cells with time, and as each cell goes 
through only one cycle during treatment there is no polyploidy, e.g., sulphanilamide. 
A third group causes chromosome breakage, e.g., nitrogen mustards. 

The cytological abnormalities produced in these experiments are similar to those 
of the second group of chemicals. A sodium-potassium phosphate buffer produced spindle 
abnormalities in meristematic cells, only in the absence of other salts. Both the 
phosphate buffer alone, and the buffer with mineral salts added, reduce the number of 
dividing cells by preventing the entry into prophase, but this effect is greater with the 
buffer solution only. The spindle mechanism was inhibited by the buffer alone, but 
not by the combined solution. Similar effects were obtained with KCl solution, which 
lias less effect on the spindle mechanism and more effect on the interphase to prophase 
stage of the cycle. Again, the addition of a culture solution with a low concentration 
of divalent as well as monovalent ions allowed the division process to continue in a 
normal manner and reduced the effect on the number of dividing cells. 

An analogous phenomenon is known in root growth in culture solutions. It is 
generally agreed thai there must be a suitable balance of the various ions in a solution, 
to maintain normal plant growth. The toxic action of a single salt such as KCl is 
reduced by the addition of quite low concentrations of a divalent salt. Antagonism, 
the influence of one ion in preventing the toxic symptoms due to another ion alone, is 
well known, but imperfectly understood. The best known example in plants, the 
antagonism of Ca or K by the other ion, presumably occurs because the presence of K 
cr Ca alone in solution results in structural disorganization of the cytoplasm. In these 
experiments the ion unbalance in the external solution affects the cell division process. 
as the cytological abnormalities produced by both the phosphate buffer and the KCl 
were neutralized at least in part by the addition of low concentrations of other salts 
to the solution. This suggests that the cytological abnormalities may be the indirect 
result of physiological changes in the cells, and are not due to the specific effect of any 
inorganic ion. This inhibition of the cell division process may be one factor which 
contributes to the inhibition of root growth in an unbalanced solution. 

Acknowledgements. 

The author wishes to thank Mr. J. B. Douglas for advice, and Professor F. V. Mercer 
and Dr. R. N. Robertson for advice and helpful criticism of the manuscript. 

■' References. 

CI.ARK, W. M., 192S. — The determination of hydrogen ions. Pp. 210. Baltimore, William.s & 

Wilkins. 
Galinskt, I., 1949. — The effect of certain pho.sphates on mitosis in Allium roots. J. Hered., 

40: 289-29.5. 
HiNDMARSH, M. M., 1951. — A critical consideration of c-mitosis with reference to the effects 

of nitrophenols. Proc. Ljnn. See. N.S.W., 7G : 158-16?,. 

, 1953. — Ph.D. Thesis. University of Sydney. 

HOAGLANDj D. R., and Broyer, T. C, 19.3(). — General nature of the process of salt accumulated 

by roots, with description of experimental methods. PI. Phi/s.. 11 : 471-507. 
Hughes,, A. F., 1950. — The mitotic cycle. London, Butterworths. 
, 1952. — Some effects of abnormal tonicity on dividing" cells in chick tissue culture. 

Quart. J. Micros. ScL, 93 : 207-219. 

, 1952. — Inhibitors and initotic physiology. S.E.B. Symposia, 6 : 256-263. 

I.BVAN, A., 1945. — Cytolog-ical reactions induced by inorganic salt solutions. 'Mature. 156 : 750- 

751. 
, 1949. — The influence on chromosomes a,nd mitosis of chemicals as studied by the 

Allium test. Hth Int. Cong. Genet. (Here.ditas suppl. 1949) : 325-337. 
, and T.IIO. .1. H., 1948a. — Chromosome fragmentation induced by phenols. Heredita.s. 

34 : 250-252. 



BY MARY M. HINDMAKSH. 231 

IjBVANj a., and Tjio^ J. H., 1948b. — Induction of chromosome fragmentation by phenols. 

Hereditas, 34 : 453-483. 
, and Wangheim, K. H., 1952. — Potassium cyanide in the Allium test. Hereditas, 

38: 207-313. 
♦Lorbnzo-Andreu, a., 1951. — Accion de varias salts alcalinas sobre la division celular en 

Allium cepa. Analas de la Estacion Exp. de Aula Dei, 2 : 174-186. 
Philpot^ J. St. L., and' Stainier^ J., 1956. — The choice of the suspension medium for rat liver 

cell nuclei. Biochem. J., 63 : 214-223. 

* Seen in abstract only. 



EXPLANATION OP PLATE VIII. 
A-B : Blocked metaphase in sodium-potassium phosphate buffer. C-E : Abnormal metaphase 
and anaphase in KCl. P : Normal cell division in phosphate buffer + culture solution. G ■ 
Normal cell division in KCl + culture solution. 



232 



THE REPRODUCTION AND EARLY LIFE HISTORY OF THE GASTROPOD 

CYMATILESTA SPENGLERI (PERRY) (PAM. CYMATIDAE). 

By D. T. Anderson, University of Sydney. 

{Communicated by Miss I. Bennett.) 

(Plate ix; 17 Text-figures.) 

[Read 29th July, 1959.] 



Synopsis. 
C. spenrjleri shows clear sexual dimorphism. Pairing occurs in the spring, and the 
female produces a characteristic egg mass. Early development through a yolky veliger to 
metamorphosis takes place within the egg mass, with escape of the young to a free-living 
benthic existence at 16 days. The pattern of reproduction and early life history is typical 
for a mesogastropod prosob ranch. 



Introduction. 
Many workers have made contributions to our knowledge of the life histories of 
prosobranch gastropods, and the details that have emerged have been clearly sum- 
marized by Lebour (1937) and Thorson (1946, 1950). The majority of investigations, 
however, have been of European and American species and the life histories of the 
tropical and subtropical marine prosobranchs, especially those of the eastern Pacific, 
remain largely unrecorded. Many of the families occurring in this region are repre- 
sented in the molluscan fauna of the eastern Australian seaboard, among them the 
Cymatidae (triton shells). The following account of the early life history of Cymatilesta 
si)engleri is the first for a member of this family. 

Sex Difference and Spawning Habits. 

C. spengleri is typical of the smaller cymatids, an average adult specimen measuring 
4-5 inches in length. Like all prosobranchs, the species is dioecious, and a degree of 
sexual dimorphism is evident in addition to the presence of the penis in the male. The 
female is slightly the larger, with a thicker and less sharply conical spire and a much 
thicker rim to the mouth of the shell. Adults of both sexes are found in considerable 
numbers near the low water mark on rocky shores in the vicinity of Sydney. Breeding 
takes place in the spring (October-November), the animals being found associated in 
pairs at this time, and egg masses are attached to shaded rock faces just above thp 
extreme low level of spring tides. The extent of the sublittoral range of the species 
and the possibility of sublittoral breeding are undecided, but an indication of a more 
than littoral distribution is given by the preponderance of fully grown adults on the 
shore. As will be seen below, the species has no planktonic larval stage and hatches 
as an almost complete miniature adult. Stages of growth leading to the adult must take 
place mainly in sublittoral waters, and it may even be that the later migration of 
adults to the shore is a breeding migration. 

Copulation has not been observed in C. spengleri, but must considerably precede 
spawning by the female, since egg laying occurs in isolation. On November 11, 1958. 
Miss I. Bennett, of the Zoology Department, University of Sydney, and myself, while 
collecting- at Long Reef, Collaroy, north of Sydney, found in addition to several egg 
masses suspected to be of this species, a female in process of spawning an identical 
egg mass. Both egg mass and parent are shown in Plate ix. 

The Egg Mass. 
When complete, the egg mass (PI. ix. A, B; Text-fig. 1) is circular in outline, cup- 
shaped, and glued to the rock by its fiat base. It consists of a spiral of conical egg 
capsules, beginning in the centre of the mass with the capsules vertically placed and 

Proceedings of the Ltxnean Society of New^ South Wales, 1959, Vol. Ixxxiv, Part 2. 



BY D. T. ANDERSON. 






continuing outwards and upwards until the capsules at the rim of the cup are 
horizontally placed. The outer surface of the mass is made up of a series of thin, 
horny, overlapping, transparent plates, with the close-packed capsules firmly attached 
to their inner surfaces. 

The female fashions the egg mass within the mouth of the shell, and the diameter 
of the two correspond. The region of greatest height of the mass is formed at the 
anterior rim of the shell, the opposite lowest region lying under the first coil of the 
visceral spire. The mode of construction of the mass is not yet clear, but the individual 
egg capsules are extruded from the mantle cavity already fully formed. The horny 
plates must therefore be produced and glued together and the capsules attached to 




— B 




Text-fig-. 1, 2. Cyinatilestd spengleri: 1. Egg- mass, natural size. 
A, horny plates of surface layer ; B, egg capsule. 



capsule, X 2. 



them by a means accessory to the genital system. Each of the several hundred egg 
capsules is a horny, transparent, flattened, blunt ended cone curving outwards at the 
base (Text-fig. 2). Within it is a clear jelly in which lie several hundred eggs, each 
creamy white in colour, but imparting to the whole egg mass a delicate pale orange 
tint. 

Methods. 

The newly spawned egg mass was maintained in a bowl of clean sea water in the 
laboratory. With constant aeration and a change of sea water every three days, the 
embryos within the >egg capsules developed steadily, feeding on yolk reserves, for 16 
days, when hatching took place. It did not prove possible to culture the hatched 
embryo further, but by this time they had almost completed metamorphosis and were 
benthic in habit. 

At intervals during the 16-day period, capsules were removed from the mass and 
slit open to release embryos for observation alive. Camera lucida drawings were made 
of specimens narcotized with 10% alcohol, magnesium chloride proving ineffective at 
normal narcotic concentration. 

ExTEKNAL Features of Development. 

The fertilized egg of C. spengleri is opaque, spherical, 120m in diameter, creamy 
white in colour, and covered by a very thin egg membrane closely adherent to the egg 
(Text-fig. 3). Before cleavage begins the membrane disappears, and the cleavage 
blastomeres are only loosely held together in the early cleavage stages. The egg is 
very yolky, large yolk globules being uniformly distributed throughout the cytoplasm. 

The first two cleavages are almost equal, but the third is very unequal, giving a 
quartette of small micromeres at the animal pole and four large yolky macromeres at 
the vegetal pole. Cleavage follows the normal spiral pattern, but all the micromeres 



234 KEPEOnUCTlON AND EARLY LIFE lIlSTOIiY OF CYMATILESTA 3PEXGLEKX, 




Text-flg-. 3-13. Cyinatilesta spengle7-i: 3. Fertilized egg-. 4. 5 hr. blastula, lateral view. 
0. 14 hr. gastrula, lateral view. 6. Early trochophore, 17 hr., dorsal view. 7. Trochophore, 30 
hr,, dorsal view. 8. Trochophore, 24 hr., ventral view. 9. Early veliger, 36 hr., dorsal view. 
10. Veliger, 2J days, dorsal view. 11. Veliger, 3i days, ventral view. 12. Veliger, 6 days, 
dorsal view. 13. Veliger, 8 days, dorsal view. 

A, yolk ; B, micromeres ; C, lateral ciliated cells ; D, stomodaeum ; E, archenteron ; F, 
oral hood ; G, shell ; H, foot ; J, yolk masses. 



BY D. T. ANDERSOK. 235 

are small, the majoi- part of the yolk residing in the four vegetal pole macromeres. A 
solid blastula with a cap of animal pole micromeres is found about live hours after 
first cleavage. 

Gastrulation proceeds wholly by epiboly, the micromeres multiplying and spreading 
over the macromeres (Text-flg. 4), at first equally on all sides, but later more rapidly 
dorsally and posteriorly, so that eventual closure of the blastopore takes place mid- 
ventrally from behind forwards, the anterior end of the blastopore remaining open as 
the future mouth. By the time gastrulation is complete the large yolky (endoderm) 
cells have begun to increase in number and the embryo shows some increase in size, 
though still retaining a spherical shape (Text-fig. 5). At this stage it is about 14 hours 
old. 

During the next three hours the embryo elongates along the antero-posterior axis 
and becomes ovoid, pointed anteriorly and blunt posteriorly (Text-fig. 6). The superficial 
layer of yolk-free cells remains very thin, while the endoderm cells are further 
increased in number and can now be seen to surround a central space, the archenteron. 
On either side at the widest part of the embryo one of the superficial cells becomes 
enlarged and projects above the general surface, showing a large clear nucleus. By 
the time the embryo is 20 hours old these cells have become ciliated, and a further 
area of short cilia has appeared at the posterior end (Text-fig. 7). At the anterior end 
multiplication of the superficial cells is beginning, giving a layer more than one cell 
thick. The embryo now corresponds to a trochophore in which, associated with the 
presence of yolk, the laryal organs are largely suppressed. 

With further development the outline of the trochophore becomes better defined, 
the anterior end becoming more pointed. The cilia on the large lateral cells disappear, 
but by 24 hours a new ciliated area is present at the anterior end (Text-fig. 8). The 
outline of a cylindrical stomodaeum can also be seen, while the archenteron now shows 
as a central cylindrical space running along the antero-posterior axis, at right angles 
to the stomodaeum. 

In the succeeding 36 hours (Text-fig. 9, 10) very little increase in size takes place, 
but the shape of the embryo changes in various ways. The main mass of the body 
containing the yolky endoderm becomes spherical and sharply demarcated from the 
anterior ciliated region, which broadens as a flattened oral hood overlying the 
stomodaeum and bearing short cilia on its convex upper (dorsal) surface. The cilia 
beat vigorously, driving a current backwards over the body, but cause no movement 
through the jelly at this stage. Neither is there any muscular movement. The 
posterior patch of cilia is displaced towards the left as a result of asymmetrical growth 
of the body, and a further patch develops laterally on the same side. The significance 
of these cilia is obscure. 

The embryo now begins to increase in length, in size of the oral hood and m 
clarity of definition of the stomodaeum, which now has a ciliated lumen (Text-fig. 11). 
The remainder of the body remains opaque and yolky, obscuring the internal changes 
taking place. The 3J-day embryo, however, shows one new surface feature in the 
presence of a small cap shell at the posterior end of the body, towards the right side, 
secreted by an ectodermal thickening, the shell gland. 

During the next three days, with little further increase in the size of the embryo, 
the shell spreads anteriorly over the dorsal surface and also downwards over the 
posterior end of the embryo as a broad cap (Text-fig. 12). The posterior and lateral 
cilia disappear, while a posterior fold of the body surface immediately underlying the 
shell marks the first origin of the mantle cavity. Mid-ventrally a transverse ciliated 
protuberance grows out, the rudiment of the future foot. The embryo is now clearly 
divided into head anterior to the shell, foot ventrally, and visceral hump covered by 
the shell. The yolk still obscures details of internal structure, but it is at this stage 
that the first muscular movement begins, a twisting of the posterior end of the visceral 
hump forwards and outwards towards the right, preliminary to torsion. 



23fi 



HEPIiODUC'lION AND EAULY LIFE UISTOKY OF CYMATILESTA 8PENGLEEI, 



By the time the embryo is eight days old torsion of tlie visceral hump is completo 
(Text-fig. 13). The region joining the visceral hump to the head and foot narrows and 
behind it the hump becomes globular and shows the differentiating gut coiling 
posteriorly, then dorsally, over it. The shell also attains a globular form, with a wide 
moutli from which the head and foot project. Withdrawal into the shell is not 
possible at this stage. The oral hood is by this time deeply convex and differentiated 
into a narrower anterior and broader posterior part, heralding the formation of the 
velum. The foot increases in size as a transverse antero-posteriorly flattened projection 
covered on the anterior face and edges by numerous short cilia. The posterior face of 
tht> foot bears a small thin operculum. Embedded in the base of the foot on either side 




Text-fig-. 14-17. Cymatilesta spengleri: 14. Veliger, 9 days, lateral view. 15. Veliger, 10 
days, ventral view. 16. Metamorphosing veliger, 16 days, ventral view. 17. Metamorphosing 
velig-er, 16 days, lateral view, showing- withdrawal into shell. 

A, oral hood : B, velum ; C, foot ; D, otocyst : E, eye ; F, stomodaeum ; G, shell ; H, 
operculum ; J, intestine : K, yolk. 

is a spherical, hollow otocyst containing a small, round, retractile otolith. The foot 
and head are now almost devoid of yolk, and the embryo appears at this stage to 
extrude yolk masses enclosed in a thin layer of cytoplasm from either side just anterioi 
to the shell margin. The reason for this is not clear. 

Growth and differentiation proceed rapidly over the next two days (Text-fig. 14, 15). 
The yolk gi-adually disappears in the visceral hump, with accompanying differentiation 
of the viscera, especially of the coiled gut, which now shows peristalsis. The shell, 
showing clear evidence of spiral form, enlarges and widens at the mouth, and a 
columella muscle faintly visible on the left side of the visceral hump in the living 
embryo is able to effect partial withdrawal of both head and foot into it. The foot 
becomes broader and its operculum larger and thicker. In the head, the anterior part 
of the oral hood becomes greatly enlarged and covered with short cilia dorsally, whih.' 
the posterior part broadens and grows ventrally towards the stomodaeum, forming a 
circular rim on which are borne the long, strongly beating cilia of the velum. The 
velar cilia beat backwards, with typical propulsive and recovery strokes, and also in a 



BY D. T. AJXDERSON. 237 

clockwise direction viewed from the anterior end, in such a way that metachroual 
waves of beating pass successively round the velar rim. The total effect of the cilia 
on the head is to drive the embryo forward through the jelly, and observation of an 
egg mass at this time gives the impression of vigorous activity. 

Within the velar rim and around the stomodaeum, rudiments of the adult head are 
now beginning to become organized, and a pair of black eyespots is especially 
conspicuous on either side of this part of the head. No obvious response to unidirect- 
tional light is detectable, however, in embryos of this age. 

The velum enlarges further on the twelfth day and projects as two lateral lobe.s 
from the sides of the head. Following this, a period of four days supervenes during 
which there are no gross changes in external form, though a gradual diminution in the 
remaining yolk in the visceral mass and an increase in muscular movement of all 
parts of the embryo indicate continuing histogenesis. The sixteenth day of develop- 
ment, however, yields the onset of a new period in the life history with retrogression 
of the "larval" organs of the embryo and hatching from the capsule. The anterioi' 
part of the oral hood is rapidly resorbed, together with the large lateral lobes of the 
velum, so that the head becomes bilobed and compact, with the two eyespots anteriorly 
and a tight velar ring around its margin (Text-flg. 16). With this change, the entire 
anterior end can now be withdrawn into the shell and the mouth of the shell closed by 
the operculum (Text-fig. 17). Locomotion is still brought about by the velar cilia, but 
these now beat not continuously but in coordinated short bursts, stopping and starting 
together. At about this time the capsule wall breaks down at the free end, apparently 
ab a result of putrefaction, and the embryos escape rapidly from the jelly into the 
surrounding sea water. They are benthic in habit from the start, the ratio between 
the power of the reduced velum and the size of the body being too low to allow 
swimming except very close to the substratum. Crawling and active feeding presumably 
soon follow. 

Discussion. 
The summary of information on the life histories of British prosobranchs given by 
Lebour (1937) provides a background against which the life history of C. spengleri can 
be viewed. It is characteristic of the mesogastropod prosobranchs that their eggs are 
laid in thick-walled protective capsules filled with a gelatinous, usually nutritive 
medium in which the eggs float. Each egg has a very thin egg membrane which 
disappears early in development, and follows a course of development through spiral 
cleavage to a well-developed veliger larva. In these respects C spengleri is typical of 
its group. Lebour clearly shows, however, and Thorson (1946) amply confirms, that 
hatching from the capsule may vary even within closely related species from the 
veliger stage, followed by a long planktotrophic life, to the fully metamorphosed stage, 
followed by a crawling benthic existence. C. spengleri, typical of species with a heavily 
yolked egg, approximates to the latter condition; but it cannot be assumed that other 
cymatids will have life histories lacking a planktonic stage. 

Acknoxoledgements. 
I should like to thank Miss I. Bennett for advice on the collection of material and 
for the photographs of Plate ix, A, B, Professor P. D. F. Murray for the photograph of 
Plate ix, C, and Mr. W. H. I. Dawbin for discussion of the paper in manuscript. 

References. 
Lhbour, M. v., 1937. — The eggs and larvae of British prosobranchs, with special reference to 

those living- in tlio planltton. Jour. mar. hiol. Ass., 22 : 105-166. 
Thorson, G., 1946. — Reproduction and larval development of Danish marine bottom 

invertebrates. Mccld. Komm. Havunderspg., Ahh. (Plankton), 4: 1-523. 
TiiORSON', G., 1950. — Reproduction and' larval ecology of marine bottom invertebrates. Biol. 

Rev., 25 : 1-45. 

EXPLANATION OF PLATE IX. 

A, B. Cymatilesta spenyleri. Female and egg mass (xi). C. Cymatilesta spengleri 
Veligfirs, 11 days. 



238 



A NEW MITE PARASITE (HARPYRHYNCHU8) FROM THE ROSELLE PARAKEET 

(TROMBIDIFORMES, ACARI). 

By R. F. Lawrence, Natal Museum, Pietermaritzburg, South Africa. 

(Communicated by Dr. A. R. Wooclhill.) 

(Two Text-figures.) 

[Read 2S)th July, 1959.] 



A new species of avian sliin parasite, Harpyrliynclnis rosellacintis, belonging to the order 
of Trombidiform mites, is described. The mite was taken from a cyst in the skin of the 
roselle parakeet and is the first Harpyrhynchus to be described from an indigenous Australian 
bird. 



Introduction. 

During 1958 Dr. Marc Andre, of the ficole pratique des Hautes Ktudes, Ministere de 
I'ifiducation Nationale, Paris, sent me a number of avian parasites from the Trouessart 
Collection; among them was a good series of mounted specimens of a new species of 
Harpyrhynclius from the roselle parakeet Platycercus eximius. This appears to be the 
first indigenous representative of the genus from Australia and as such is of some 
interest, also in view of the fact that in general it differs very little from the various 
species of the genus found in Europe and Africa while resembling two of these in some 
detail. The slides were labelled Sarcoiborus (a synonym for Harpyrhynclius) crista- 
yalli Berlese & Trouessart, which is obviously a slip, as the Harpyrhynchus crista-galU 
described by these two authors (1889, p. 139) from the African speckled Coly, Colius 
striatus, has a quite different and very characteristic appearance; it has recently been 
figured and redescribed by Lawrence (1959, fig. 1). 

I am greatly indebted to Dr. Marc Andre for allowing me to describe this new 
form. The Holotype slide and the bulk of the material have been returned to the 
collection of which Dr. Andre is in charge, while a paratype slide will be deposited in 
each of the following institutions: The Australian Museum, Sydney, Australia and the 
Natal Museum, Pietermaritzburg, South Africa. 

Family HARPYRHYNCHIDAE Dubinin. 

Genus Hakpyriiynchus Megnin. 

Harpyrhynchus rosellacinus, n. sp. (Figs. 1, 2). 

Material: A mounted series of 33 adult females and 11 larvae on 13 slides, taken 
fiom a cyst in the skin of the parakeet Platycercus eximius at Sydney (no further 
data), in the collection Trouessart. 

Holotype: One slide with 4 adult females. Paratypes: Three slides with 5, 2 and 1 
adult females respectively. 

Dorsal surface as in Figure la. dorsal shield clearly defined, a little longer than 
wide, with two setae situated just within its anterior border, these a little longer than 
the pair laterally and a little posteriorly to the peritremes, and subequal to the lateral 
setae between the insertion of legs I and II. Lateral margin between legs II and III 
with two similar setae situated close to each other, the one on the edge of the body, the 
other (a little in advance of it) inserted on the dorsal surface. All these setae with 
fine accessory hairlike serrations, a little stronger in the two last-named setae, all 
the remaining setae of the body and legs simple. 

Ventral surface as in Figure 2 (of a larger paratype J), the setae arranged as iu 
Figure 2, rising from large circular disc-like areas; posterior margin of body in the 
middle with a short slender seta on each side of the anal opening, very near the margiii 
but definitely inserted on the ventral surface. 

PROCEEDiNGe OF THE LiNNEAN SOCIETY OF New SorTH Wales, 1959, Vol. Ixxxiv, Part 2. 



BY R. F. LA'WRENCE. 



239 



Pedipalps as in Figure Ic, seen from above, enlarged, from below as in Figure 2; 
dorsal surface with 3 large serrate hairs, the anterior one thickest and with about 18 
tooth-like serrations arranged in a regular row on its dorsal margin, these modified 
hairs situated well in anterior third of the dorsal swelling; in addition a stout simple 




Fig. 1. Harpyrhynchus rosellacinus, n. sp., ?. a, dorsal surface (holotype) ; ft, leg- I 
in dorsal view $ (paratype) ; c, pedipalp in d'orsal view enlarged (paratype). 



seta rising from near the middle point of the pedipalp and laterally to the centre, this 
seta unusually long, easily surpassing the peritremal openings on each side; basal 
segment of pedipalp with 2 conspicuous setae on its ventral surface (Figure 2). 

Legs: Leg III with 2 or 3, IV with only 2 very short rounded segments; III with 
4 or 5 long terminal setae (5 probably the correct number) of which 3 are considerably 
thicker and longer than the two others; leg IV with 3 terminal setae, 2 being long and 
thick (Fig. lo, 2). The thickest seta of tarsus III a little stouter than the thickest of 



240' 



A NEW MITE PAEASITE FROM THE EOSELI.IC PAKAKEET. 



tarsus IV, the length of these two setae subequal and about equal to the total length of 
body. In Figure la the anterior legs of the holotype are bent over and downwards so 
that their apices are obscured; Figure lb of another adult female shows leg I seen from 
above in full extension. 

Dimensions: Length and width of holotype $ (mouth-parts included), 310^ and 264|U 
respectively; of a larger paratype J, 370/x and 278^. 

Larim.: Round, wider than long, with only three pairs of well-developed legs, the 
two anterior ones with relatively much longer setae than in the adult, but with similar 




Fig. 2. HaypyrhyncUus roaellacinus n. sp. Ventral surface of a paratype 



claws and other terminal structures; leg III with 2 or 3 short, rounded segments, the 
apical one with three setae of different lengths, the second longest much exceeding the 
longest setae of the anterior legs, the longest about li times the total body length. 

Dorsal plate well defined, pedipalp well developed, with 3 modified hairs dorsally 
similar to those of adult, but no long simple seta posterior to these. 

Dimensions: Total length and width of body, 120/i and 121p. respectively; longest 
seta of leg III, 209^. 

No nymphs or male specimens are represented in the material before me. A number 
of encapsuled eggs are mounted with adult females on some of the slides; long setae 
and appendages at various stages of development can be seen through the capsules of the 
eggs which seem to be embedded in a structureless colloidal substance, possibly the 
contents of the cyst. 

Affinities: The parasite resembles two species of Harpy rliynclius more closely than 
any others, nidulans Megnin and tracheatus Fritsch. In having the dorsal setae pro 
vided with fine, almost invisible serrations, it resembles nidulans rather than 
tracheatus. but differs from it in the modified dorsal hairs of the palp being much 



BY R. F. LAM'RENCE. 241 

more heterogeneous in size and provided with stronger, more numerous and more 
regular serraitions; the long smooth seta behind these hairs is either absent in nidulans 
or situated on the lateral surface; furthermore, the arrangement of the ventral setae 
of the abdomen is quite different in the tv^^o species and the number of long terminal 
setae on legs III and IV is also different, these being respectively 6-8 and 5-6 according 
to the illustrations of Fritsch for nidulans, 1954, fig. 1 (5-7 and 4 respectively according 
to the figures given by Dubinin, 1957, p. 97-99, figs. 23 and 25, for the same species). 

From tracheatus it differs in the smooth dorsal seta of the pedipalp enlargement 
being much longer and situated further posteriorly, reaching beyond the peritreme, 
while in tracheatus it falls far short of it. The dorsal shield is comparatively wider 
and the setal pattern of the ventral surface is different from that of tracheatus, but 
the number of long terminal setae for legs III and IV is the same. 

Taking all these characters into consideration, rosellacinus appears to resemble 
ti'acheatus more closely than any of the hitherto described forms, but is distinct from 
it; tracheatus was taken by Fritsch from a host belonging to a very different order of 
birds since it was found on the common buzzard of Europe, Buteo buteo. 

References. 

Bbrlbse, a., and Trouessart, E., 1889. — Diagncses d'Acariens nouveaux ou peu connu.s. Bull. 
Bihl. Sci. de VOuest, Ann. 2: 121-143. 

Dubinin, V. B., 1957. — A new classification of the suborders Cheletoidea W.Dub. and 
Demodicoidea W.Dub. (Acariformes, Trombidiformes). Parasitol. Jour., 17 : 71-136 
(Moscow). 

Fritsch, W., 1954. — Die Milbengattung- Harpyrhynchus Megnin 1878 (Subordo Trombidi- 
formes, Fam. Myobiidae Megnin). Zool. Anz., 152, (H.7/8), 177-198. 

Lawrence_, R. F., 1959. — New mite parasites of African Birds (Myobiidae, Cheyletidae). 
Parasitology, Cambridge (in press). 



242 



FLORAL STRUCTURE AND ANATOMY IN THE FAMILY GOODENIACEAE 

DUMORT. 

By R. C. Carolin, University of Sydney. 

(Forty-five Text-figures.) 

[Read 29th July, 1959.] 



Synoi)sis. 
The floral anatoir,y of most genera in this family has been examined. It is concluded 
that the ovary of the Goodeniaceae is 4-carpellary, although various reductions and fusions 
obscure this. The ovary is fairly constant in form ; a basal 2-locular zone, a medial unilooilar 
zone and an uppermost 2-locular zone, the latter two being sterile. The relative, sizes of these 
zones give the appearance of partial or complete 1-locular or 2-locular conditions. Solidifica- 
tion of one loculus may also give an apparently 1-locular condition. The inferior condition of 
the ovary may, in general, be considered to have arisen by fairly superficial mass growth, 
i.e , by fusion of the outer floral parts to the ovary. The evolution within the family is 
considered in the light of these investigations. It is noted that floral form alone hardly 
indicates a Campanulaceous origin, , 

Introduction. 

The use of vascular patterns as a guide to structure lias been a controversial 
technique. The crux of the problem appears to be the lack of knowledge of the 
fundamental cause of the initiation of a vascular strand. Puri (1951), Douglas (1944, 
1957), and Fames and Macdaniels (1947) have summarized the arguments in favour 
of the conservative nature of vascular patterns. It is not intended to reiterate these 
arguments, suffice it to say that they appear reasonable. These patterns are controlled 
by the genotype of the plant and, because of their relative constancy, it can be assumed 
that the genetic systems responsible for them are well "buffered" against interference 
from the environment. The numerous cases of vascular strands pointing to the assumed 
positions of aborted organs implies that this genetic system is "conservative", i.e., new 
floral forms are superimposed on the previous vascular structures. Thus the vascular 
pattern of the flower can be used as a guide to changes in genotype, i.e., phylogeny. 

Colozza (1907, 1908) has described the vegetative anatomy and Brough (1927) 
examined the ontogeny of the flower of a single species. Saunders (1939) gives some 
data on a few genera; this is not complete and the interpretation is coloured by the 
theory of carpel polymorphism. This theory is considered to be an unnecessary com- 
plication. 

Material and Methods. 

Mature flowers and often young fruits were embedded in wax in the usual manner 
using chloroform as the wax solvent. Both transverse and longitudinal sections were 
cut at a thickness of 15^. The sections were stained with crystal violet and counter- 
stained with erythrosin. In most cases fresh material was fixed in formalin-acetic- 
alcohol. In a few cases herbarium specimens were soaked in warm detergent before 
embedding in wax. The results are presented as diagrams in the main; only a few 
of the drawings of the serial sections have been reproduced here. In view of the fact 
that a number of species are known to show variability in their floral vascular patterns 
(Hall, 1956) a number of specimens (between 5 and 15) were examined for each species. 
The nomenclature of the vascular strands is that used by Eames and Macdaniels (1947). 

Observations. 
The carpels appear to be considerably modified and the loculi walls are probably 
formed by several carpels. The term pseudocarpel is therefore used in this presentation 
of the results to include the walls of each loculus and the corresponding part of the 
septum (see below). 

Proceedings of the Linnean Society of New South Wales, 1959, Vol. Ixxxiv, Part 2. 



BY E. C. CAKOLIN. 



243 



The point of reference for the genus may be taken as Velleia with a superior ovary 
and semi-epigynous stamens. This genus will be described rather more fully and the 
others compared with it. 

Velleia Smith. 

There are two main floral types within the genus, those with five sepals and those 
with three. 

V. paradoxa R.Br, has five sepals. The first vascular strands which diverge from 
the central stele of the receptacle are on the sepal radii; almost immediately above 
these the strands on the petal radii also diverge and these latter almost immediately 
branch into three. The two lateral members of each group unite with the strands on 
the sepal radii, one on either side, whilst the central strand continues on into the petals 
without further change. Five further bundles diverge from the central stele on the 




Text-figures 1-3. 
1. Velleia lyrata, t.s. of lower flower. 2. Goodenia scapigera, part of pseudo-ovary. 
3. Goodenia decurrens, t.s. of receptacle showing vascular strands before their fusion. 



sepal radii (presumably the staminal bundles) which unite with those already on 
these radii (Fig. 3). The vascular elements remaining in the central stele resolve into 
four or five strands which further unite into two strands diverging outwards and a 
single strand in the centre. Those diverging outwards (pseudocarpellary dorsal 
bundles) are on the post.-ant. axis, the posterior one uniting with the posterior strand 
on the sepal radius and the anterior one doubling backwards into the spur. The outer 
floral whorls themselves are fused to the ovary at the base; the sepals become free 
first, except for the posterior one which remains united to the ovary for some consider- 
able distance. In the lower quarter of the flower the petal lateral, staminal and post- 
pseudocarpellary dorsal bundles all diverge from the appropriate strand on the sepal 
radii, which latter passes into the sepal itself where it branches into three. The 
petals diverge somewhat higher up as a complete tube (Fig. 1, 2) into which the petal 
strands pass. Still higher the stamens diverge, also supplied by their own vascular 

GG 



244 



FLORAL STRUCTUKE AND ANATOMY IN THE FAMILY GOODENIACEAE, 




Text-figures 4-10. 

4. Goodenia decurrens, t.s. of lower part of pseudo-ovary. 5. Anthotium rubriflorum, t.s. 
of pseudo-ovary. 6. Dampiera stricta, t.s. of pseudo-ovary at level of ovular insertion. 7. 
Leschenaultia biloha, t.s. of pseudo-ovary. 8. Leschenaultia hiloba, t.s. of ovary dome. 
9. Goodenia decurrens, t.s. of upper part of pseudo-ovary. 10. Scaevola albida, t.s. of ovary 
dome. 

5, sepal main ; SI, sepal lateral ; P, petal main ; PI, petal lateral ; St, stamen ; Cd, Ps-carp. 
dorsal ; CI, Ps-carp. lateral ; V, Placental ; Sp, spur or space. 



BY K. C. CAROLIN. 245 

supply (Fig. 3, 31). This fusion of the various vascular supplies and their subsequent 
separation from each other is illustrated in Figure 40, although in this case the 
separation occurs lower in the flower than in Goodenia. The anterior petal is drawn 
out into a spur whilst the anterior wall of the ovary is nectiferous. The remnants 
of the central stele pass into the placentae as an irregular bundle. Very low down two 
bundles pass out from it along the septum and take up a position at the junction of 
septum and ovary wall (ps.-carp. lat. bundles). The central bundle progresses up into 
the septum and branches of it supply the ovules. In addition two further sterile 
bundles pass out on either side towards the ovary wall and there fuse with the pseudo- 
carpellary lateral bundles (see Fig. 4). The remaining vascular tissues are entirely 
used up in supplying the uppermost ovules. 

The ovary represents the standard Goodeniaceous type, the general form of which 
is repeated throughout the family. In the lowest part the ovary is divided by a single 
septum with an axile placenta (Fig. 1, 4). About halfway up the ovary a split appears 
in the axis, thus giving a semblance to parietal placentation (see Fig. 9). Very few 
ovules, or none at all, are inserted in this region; they are mainly found in the lower- 
most zone. Just below the ovary summit the inturned edges of the pseudocarpels meet 
once again, regenerating the apparently bilocular condition. Thus there is a lower- 
most "bilocular" zone, a medial "unilocular" zone and an uppermost "bilocular" zone 
(Fig. 31). 

The style is supplied by the pseudocarpellary dorsal and lateral bundles which do 
not, however, pass into the style unchanged. They bifurcate as shown in Figure 11, 
and these branches fuse as indicated to give the condition of four bundles as shown 
in Figures 12 to 15. These four bundles are arranged over the loculi, none over the 
junction of septum and ovary wall as would be expected if the loculi represented 
individual carpels. Immediately below the indusium these bundles undergo a series of 
bifurcations to form a ring of vascular tissue. 

y. lyrata R.Br. — This species has three sepals, but otherwise is very similar to the 
previous species. The lateral sepals are united to their anterior lateral partners and 
the vascular supply of these compound sepals still retains some degree of separateness 
(Fig. 1). 

V. montana Hook. f. — Essentially the same as the preceding species. 

Symphyobasis. 

iS. macroplectra (F. Muell.) Krause. — Unfortunately no material of this interesting 
monospecific genus was available. It would appear to be similar to Yelleia except that 
the corolla and stamens are virtually epigynous, whereas the sepals are almost entirely 
hypogynous, i.e., the ovary is superior. 

Goodenia. 

G. decurrens R.Br. — The basic plan is similar to that found in Velleia paradoxa. 
The sepals and other outer floral whorls separate from the ovary wall at a much higher 
level and the condensation of the vascular system is correspondingly increased (Fig. 
4, 9, 32). The ribs on the floral tube each contain a bundle which divides at the point 
at which the rib diverges from the ovary wall as a sepal, into six as shown in Figure 40, 
although the bundle supplying the stamens may diverge soJtiewhat lower (Fig. 4). 
These bundles are: sepal main, two sepal lateral, petal laterals and staminal. In 
addition the posterior bundle on the sepal radius gives rise to one of the pseudo- 
carpellary dorsal bundles (Fig. 40). These ribs have a different texture, and in 
some other species a different indumentum, from the grooves between them. The 
texture and the indumentum correspond exactly to that of the sepals which are formed 
by the separation of these ribs from the pseudo-ovary wall. The grooves are supplied 
with single bundles which pass directly into the petals as petal main bundles. The 
anterior pseudocarpellary dorsal bundle diverges from the bundle on the anterior petal 
radius at a low level in the pseudo-ovary wall and is separated from the ovary wall 



246 



FLORAL STRUCTURE AND ANATOMY IN THE FAMILY GOODENIACEAE, 



proper by a spur or pocket (Fig. 4, 9, 32). In other respects the vascular pattern is 
similar to Velleia (also see Fig. 14, 40). 

G. ovata Sm. and G. bellidifolia Sm. showed only minor differences. 

G. heterophylla Sm. — In this species the sepals do not diverge from the pseudo- 
ovary wall until very close to the summit of the ovary and there is a corresponding 
condensation of the vascular tissue. 




13 



15 




• • 



• • 



Text-flgures 11-30. 
11, 12. Derivation of stylar vasculation in 4-strand' styles. 13. Velleia paradoxa, style. 
14. Goodenia hellidifoUa, style. 15. Goodenia pusilliflora, style. 16, 17. Derivation of stylar 
vasculation in two strand styles (except Brunonia). 18. Dampiera stricta, style. 19. Scaevola 
hookeri, style. 20. Anthotium ruhrifiorum, style. 21, 22, 23. Stylar vasculation of Scaevola 
ramosissima. 24, 25, 26. Stylar vasculation of Leschenaultia. 27, 28, 29, 30. Vasculation of 
stylar branches in Calogyne pilosa. Varying levels in style; 27 lowest, 30 highest. 



BT R. C. CAKOLIN. 247 

G. pusilliflora F. Muell. — Although agreeing with G. decurrens in most respects, the 
only specimens of this species which were examined had 2-fld styles. This branching 
reached almost to the base of the style and the cleft was at right angles to the ovary 
septum, each branch containing two vascular strands (Fig. 15) which have been 
formed in the same way as in G. decurrens. 

G. paniculata Sm. — This species is a member of the Section Pluriseratim Benth. or 
Amphichila DC. The floral construction is similar to that of the other species described 
except that the ovules are inserted in a number of rows on the placentae instead of in 
two rows. 

G. scapigera R.Br. — This species is a member of the section Monochila Benth. Once 
again it differs from the other species examined in minor details only, the most 
interesting being the attachment of the ribs (sepals) to the pseudo-ovary wall (Fig. 2). 
This attachment is by a narrow strip of tissue only; the whole inner surface of the 
sepaline rib is not adherent to the rest of the pseudo-ovary wall as in the previous 
species. 

Calogyne. 

C. berardiana (Gaudich.) F. Muell. — The floral structure of this species resembles 
that of the Goodenias very closely. The distinctive character is considered to be the 
2-fid style, the structure of which is exactly similar to that of G. pusilliflora (see above). 

C. pilosa R.Br. — This species differs from the Goodenias only in the presence of a 
3-fid style. This branching occurs quite close to the indusium, the vascular supply of 
the branches being supplied by branches from the four standard bundles in the manner 
indicated in Figures 27-30. From these diagrams it is obvious that the stylar condition 
bears little relationship to that found in the previous species or to the carpellary 
condition of the ovary. 

Sellieka. 

S. radicans Cav. — This genus has essentially the same floral structure as that found 
in the Goodenias. The sepaline ribs, however, are not evident as such and the pseudo- 
carpellary dorsal bundles separate from the corresponding outer bundles (see Figure 40) 
very low in the pseudo-ovary wall. The spur is reduced to a small anterior pocket. 

Vereeauxia. 

V. reinivardtii (de Vriese) Benth. The structure of the pseudo-ovary wall is 
essentially the same as that of the Goodenias (Fig. 33) and the derivation of the 
vascular supply of the separate organs is the same (see Fig. 40). The main difference 
is to be found in the ovary itself which is usually described as 1-locular with a solitary 
ovule. In fact the standard ovary form is present, but the lowermost "bilocular" zone 
and the uppermost "bilocular" zone are very small. The ovule is inserted in the 
posterior position on the short, lower septum. The sterile septal bundles do not occur, 
but otherwise the vascular pattern is similar to that found in the Goodenias (Fig. 33, 
45). 

Scaevola. 

Three main types of ovary are found in this genus as described by Krause (1912) : 
(i) "bilocular" with one ovule per loculus; (ii) "unilocular" with two ovules in the 
loculus; (iii) "unilocular" with one ovule in the loculus. 

8. alhida Sm. — This species has type (ii) ovary, but once again it is found that the 
structure is really that of the generalized type as found in the previous genera. The 
basal "bilocular" zone is considerably reduced and one ovule is inserted on either side 
of the short septum. Immediately beneath the style the "bilocular" condition is 
regenerated. Ten bundles diverge from the central receptacular stele almost at the 
same time. These pass outwards into the pseudo-ovary wall unaltered. There are no 
ribs on the outside of this wall. Five of them on the sepal radii behave in the same 
way as those in the same position in Goodenia, the other five on the petal radii likewise 



248 



FLORAL STRUCTURE AND ANATOMY IN THE FAMILY GOODENIACEAE, 



(Fig. 40, 10). There is, however, no spur and the anterior pseudocarpellary bundle 
diverges from the corresponding petal bundle at the summit of the ovary and not before 
(see Fig. 34). After the ten main bundles have diverged from the central stele the 
remnants resolve into a single bundle with irregular patches of phloem around a central 
xylem core. Below the septum, i.e., in the receptacle, this bundle branches into three, 
two passing outwards to the junction of septum and ovary wall (pseudocarpellary 
laterals) and the central one continuing into the placentae and supplying the ovules 
(Fig. 34). 




Text-fig'ures 31-39. Longitudinal diagrammatic sections of pseudo-ovary. 
31, Velleia ; 32, Goodenia ; 33, Verreauxia; 34, Scaevola (spinescens) : 35. Diaspasis ; 36, 
Leschenaultia ; 37, Anthotium ; 38, Dampiera ; 39, Brunonia. 



8. spinescens R.Br. — This species has type (i) ovary. Apart from the presence of 
an unvasculated septum extending almost to the top of the ovary there is little difference 
between this and the preceding species. The basal "bilocular" zone is thus very large 
whilst the medial "unilocular" zone is very small (Fig. 34, 45). 

8. hookeri F. Muell. — Similar to the previous species, but the style contains two 
vascular bundles derived as shown in Figures 16, 17 and 18. 

8. ramosissima (Sm. ) Krause. — This species displays a number of differences from 
the patterns described for the previous ones. Generally the structure is the same as that 
found in S. spinescens, but the derivation of some of the vascular bundles at the ovary 



BT K. C. CAKOLIN. 



249 



dome is different. The sepal lateral bundles actually diverge from the bundles on the 
petal radii, a condition found nowhere else in the family so far. The pseudocarpellary 
dorsal bundles are very weak and when they reach the base of the style they do not 
bifurcate; the lateral bundles do, however, and the resultant bundles remain free. This 
produces a six-strand condition in the style of this species (Fig. 21-23). 




Text-figures 40-42. Derivation of vascular bundles at the ovary dome. 
40. Goodenia group. 41. Dampiera group (in Anthotmni the ps.-carp. lateral are united to 

a sepal bundle). 42. Leschenaultia. //// Sepal Traces, ".■." Petal Traces, Stamen 

Traces, • Carpel Traces. 

S. fasciculata Benth. and S. helmsii Pritzel. — These species have type (iii) ovary. 
Their floral structure is more comparable to that of Verreauxia. Not only is the structure 
cf the ovary the same, but a spur is present containing the nectary which is not situated 
on the top of the ovary as in the other Scaevolas. 



DiASPASIS. 

D. fllifolia R.Br. — The only species described to date in this genus. The floral 
structure is similar to that of S. albida. No ribs are present on the surface of the 
pseudo-ovary and there is no pocket or spur containing a nectary. The stylar supply 
is derived in the same way as that of S. hookeri. The anterior pseudocarpellary dorsal 
bundle is free for the greater part of the pseudo-ovary wall (Fig. 35). 



250' ■ floral stkuctuke and anatomy in the family goodeniaceae, 

Leschenatjltia. 

L. tiloha Lindl. — Externally the pseudo-ovary has four ribs, one of which is larger 
than the others and is shallowly grooved. Ten bundles diverge from the receptacular 
stele almost at the same level, three of these passing into the smaller ribs, four into 
the grooves between them and three into the large rib. This large rib appears to be the 
result of the confluence of two ribs and the groove between. If this is accepted the 
derivation of the vascular supply of the various floral organs at the ovary summit is 
similar to that found in the preceding species with the exception of the stylar supply 
(^Fig. 7, 8, 36, 42). The bundles on the sepal radii very often become split into two 
(Fig. 7); this division is closed up again before the divergence- of the bundles at the 
ovary dome except for the pseudocarpellary dorsals. These diverge from the posterior 
and an anterior lateral bundle on sepal radii, as double structures (Fig. 7, 42). They 
remain separate, passing on into the style as distinct bundles. The bifurcations of the 
lateral bundles provide eight separate strands in the style (Fig. 24-26). The remains 
of the receptacular stele resolve into two bundles which pass laterally and remain at 
the junction of the septum and ovary wall. These are the placental supply (Fig. 8) 
and apparently also represent the pseudocarpellary laterals of the other genera as they 
continue on above the level of ovular insertion and bifurcate at the base of the style 
(see above). The ovary does not quite correspond to the generalized form. The 
"tailocular" zone is well developed and the ovules are inserted in it; just beneath the 
style the septum splits and a "unilocular" condition is found, which continues on into 
the style, and there is no uppermost "bilocular" zone. The style is hollow and even 
open to the exterior via the indusium. Another very peculiar condition is evident in 
the pseudo-ovary wall. A definite space is present opposite the four bundles on petal 
radii contained within the grooves. This space separates the tissue surrounding the 
bundle from the (?) ovary wall proper. The space does not communicate with the 
exterior at all as it is closed over above; there is no question of its being nectiferous 
(Fig. 7, 8, 36). 

L. linearioides DC, L. divaricata F. Muell., and L. formosa R.Br, are essentially 
the same as the preceding species. 

Anthotium. 

A. ruWiflorum F. Muell. — The derivation of the Vascular supply in the receptacle 
is similar to that found in the Goodenias. The placental supply is a single strand which 
provides pseudocarpellary laterals whilst still in the receptacle. There are five ribs on 
the wall of the pseudo-ovary and three deep spurs on the anterior petal radii (Fig. 5) 
and two much smaller ones on the posterior petal radii. The derivation of the vascular 
supplies from the bundles of the pseudo-ovary wall is similar to that of Leschenaultia. 
The main difference is to be found in the ovary supply. One pseudocarpellary dorsal 
diverges from the posterior sepal bundle close to the ovary summit, the other diverges 
from one of the anterior-lateral sepal bundles low in the pseudo-ovary wall and for the 
greater length of this latter it is free (Fig. 7). The staminal bundles also diverge from 
the corresponding sepal bundles, low in the pseudo-ovary wall. The two pseudocarpellary 
dorsals fuse with the bifurcations of the pseudocarpellary laterals to produce a two- 
strand condition in the style (Fig. 16, 17, 20). It should also be noted that the bundles 
on the sepal radii in the pseudo-ovary wall do not split, nor are the pseudocarpellary 
dorsals double as in Leschenaultia. 

Dampieea. 

With regard to floral structure there are two main types within this genus: (1) 
"bilocular" ovary with one ovule per loculus, (ii) "unilocular" ovary with a solitary 
ovule. Unfortunately no satisfactory material of the former was obtainable. 

D. stricta R.Br. — This species is a member of the latter group. In spite of the 
term "unilocular" usually applied to this ovary it shows the generalized form, although 
considerably modified basally. The posterior pseudocarpel is solid (Fig. 6, 38, 45) for 
most of the basal zone. Immediately above the insertion of the single, anterior ovule. 



BY R. C. CAROLIN. 251 

this pseudocarpel becomes hollow and then the septum aborts, leaving the "unilocular" 
condition. Just beneath the style the "bilocular" condition is regenerated. There are 
five ribs on the wall of the pseudo-ovary, each rib corresponding to a sepal. Within 
these ribs, moreover, there are three vascular bundles, a main one and two lateral ones 
which are derived from the former low down. The main one divides at the ovary 
dome to give rise to the staminal strand and two petal laterals each. The posterior and 
one of the anterior lateral ones further give rise to the two pseudocarpellary dorsals 
(Fig. 41, 45). There are small nectiferous pockets opposite the grooves, which latter 
also contain a bundle outside the pocket, which passes directly into the petal. Brough 
is in error when he states that the nectaries are opposite the stamens. The derivation 
of the main bundles of the pseudo-ovary wall within the receptacle is similar to that 
seen in Scaevola. The remnants of this stele resolve into a single strand also as in 
Scaevola which passes upwards into the placenta. At the point of ovular insertion this 
bundle divides into three, one bundle to the ovule and the other two passing outwards 
to fuse with the posterior lateral bundles on the sepal radii (Fig. 6, 45). These, 
presumably, are the pseudocarpellary laterals. Evidence is provided for this by the 
distribution of vascular bundles to the stylar base; in addition to the pseudocarpellary 
bundles mentioned above the two posterior-lateral sepal bundles distribute a strand 
each to the base of the style. These strands bifurcate and fuse with the other two as 
indicated in Figures 16, 17, 41, giving a two-strand structure to the style (Fig. 20). 

The insertion of the ovule and abortion of the solid pseudocarpel take place so 
close to the base that an impression of basal insertion is conveyed. 

D. spicigera Benth. — Generally very similar to the preceding species, but the solid 
pseudocarpel is continued further up the wall of the ovary conveying the impression of 
lateral insertion. This is characteristic of the section Linschotenia. 

D. purpurea R.Br. — Similar to D. stricta. The petal lateral bundles are free for the 
greater length of the pseudo-ovary wall and lie beside the sepal laterals in the rather 
obscure sepaline ribs. 

Brunonia. 

B. australis R.Br. — This species has a superior ovary and hypogynous, epipetalous 
stamens. Sepal, petal and stamen strands are all derived separately from the 
receptacular stele and do not fuse. The carpellary supply consists of two dorsals which 
diverge next, anteriorly and posteriorly, and the single strand resolution of the 
remnants in the septum. This latter passes directly into the ovule. The style is 
supplied directly by the dorsal strands without further division or fusion. The style 
is therefore two-stranded, although different in derivation from any of the other two- 
strand styles found in this family. The ovary shows a basal "bilocular" zone with the 
solitary ovule inserted on the short septum. There is no uppermost "bilocular" zone 
and most of the ovary is aseptate (Fig. 39, 45). 

Discussion and Conclusions. 

These results are summarized in diagrammatic form in Figures 31-42 and 44 
which can be used as a basis for discussion. 

The structure of the individual floral whorls follows the general pattern found in 
the Angiosperms with the exception of the ovary. Sepals and petals are, in the more 
primitive forms, supplied by a single bundle which divides into three, one main and 
two laterals. This is somewhat obscured, due to vascular condensation in the more 
advanced genera, but is clear in Velleia. In almost all genera examined the ovary can 
be divided into the three zones mentioned under Velleia. It is in vasculation of the 
ovary that a rather peculiar situation is found. The presence of sterile strands in the 
placental septum has to be explained, and the fact that the stylar vascular supply is not 
in the position that one would expect if the locules corresponded to carpels is also 
anomalous. Nevertheless, the situation can be explained in terms of the classical theory 
provided that the locules are considered to be derived from two carpels and not one. 



252 



FLORAL STEUCTUKE AND ANATOMY IN THE FAMILY GOODENIACEAE, 



The ancestral form of the Goodeniaceae is postulated to have had four carpels in 
the four lateral positions, each with five vascular strands, a dorsal, two laterals and 
two ventrals (Fig. 42). Concomitant with the abortion of the septa on the anterior- 
posterior axis there have been fusions of the vascular bundles as shown in Figure 43. 
This gives the four-strand condition of the Velleia ovary with a central placental supply; 
the two bundles from the original lateral septa, however, are sterile and diverge as 
the sterile septal bundles (Fig. .44). At the base of the style the vascular supply 
rearranges so that the four ancestral carpels" are represented by one single strand each 
(Fig. 11, 12) in some forms or by varying numbers of strands in others (Fig. 21-25). 
Thus it is postulated that the pseudocarpellary dorsal bundles are, in fact, double or 
quadruple bundles, a conclusion supported by their bifurcation at the base of the style 





.-a. 



Text-figure 43. 4-locuIar ancestral ovary showing fusions of vascular bundles (in the 
phylogenic sense) to produce the Goodeniaceous ovary. 

Text-figure 44. Ovary of Goodeniaceae, basic type. 

and possibly by their twinned nature in Leschenaultia (although in this latter case it 
may be simply a reflection of the double nature of their parent bundles). It is con- 
sidered that the ovary of this family has achieved a bilocular condition by fusion of 
carpels two by two and subsequent condensation of their vasculations rather than by 
simple abortion. Further development within the family has occurred by the acquisition 
of a virtually unilocular condition (along at least two different lines), reduction in 
ovular number and further condensation in the vasculation. 

Before considering the relationships of the various genera it may be profitable to 
consider the question of the inferior ovary in the family. This question really involves 
much more than can be dealt with satisfactorily here. Puri (1952) has thrown some 
doubt on the usefulness of the vascular pattern in determining the nature of the 
pseudo-ovary wall. Douglas has attempted to counter this, but does not appear to hav^, 
invalidated Puri's arguments entirely. Both of these workers have useful, although 
different, ways of looking at the same problem. It seems that the ovary of the 
Goodeniaceae becomes inferior by virtue of fusion of the outer floral whorls to the 
ovary wall (in the phylogenetic sense) which is here defined as due to relatively 
superficial mass growth in the ontogenetic sense. The similarity of the structure of 
the ribs on the pseudo-ovary wall and the sepals which continue them has been remarked 
upon in connection with Goodenia. The resemblance is even more striking in the case 
of Dampiera stricta where both ribs and sepals contain many sclereids, but the petals 
and the receptacle do not, and the ribs contain the same vascular pattern as the sepals. 
The presence of a single spur in those genera allied to Goodenia and several spurs in 
Anthotium also indicate that superficial zonal growth only is involved, otherwise the 
position of the spur would be closer to the ovary summit and one could probably expect 
some inversion of vasculation in this region. Likewise the spaces in the pseudo-ovary 
wall of Leschenaultia are most easily interpreted as incomplete fusion of the petals to 
the ovary wall, possibly a number of closed spurs. The ribs of the pseudo-ovary of 
Goodenia scapigera likewise are most easily interpreted as incomplete fusion of the 
se])als to the inner floral whorls. Lastly there is the phylogenetic sequence from Velleia 
through Symphyobasis to Goodenia which is most easily interpreted as increasing 



Br R. C. CAEOLIN. 253 

fusion of floral whorls or increasingly early onset of coalescence of growth areas in 
ontogeny. There is no internal evidence with respect to Selliera, Scaevola or Diaspasis, 
but it is reasonable to suggest that the inferior ovary in the case of this family has 
arisen by superficial zonal growth, although there is the possibility that this growth 
zone may move into the inner parts of the receptacle as evolution progresses. The 
difference between the two theories postulated to explain the floral tube and the pseudo- 
ovary wall becomes smaller when it is considered in this light. 

Four main lines of development in the family may be distinguished. Brunonia 
stands somewhat apart and, indeed, has been elevated into a monotypic family. The 
hiypogynous position of the stamens and the very reduced nature of the ovary, separate 
it from other genera. The former condition may be looked upon as primitive, whereas 
the latter may be considerably advanced. The different nature of the style vasculation 
has been remarked upon above. Primulaceae and Gentianaceae have both been 
mentioned as possible ancestors for this genus, but the generalized structure conforms 
fairly well with that of the rest of the family. 

Yelleia shows a certain amount of fusion between floral whorls at the base of the 
flower, the inner floral parts being united to the ovary to a higher level than the outer 
ones. This fusion is carried still higher in the flower in Symphyobasis, and in Qoodenia 
the sepals are fused often almost to the ovary dome. Calogyne is a differentiate from 
Goodenia with divided indusia. Selliera is also a differentiate from Goodenia, dis- 
tinguished on the fruit and corolla. Pentaptilon and Catosperma may be representatives 
of this line with pendulous ovules. In some species of Goodenia the number of ovules 
is reduced to two and in Verreauxia to one with a concomitant increase in the size of 
the "unilocular" medial zone of the ovary (see Fig. 31-33, 45). The presence and 
position of the single spur and the derivation of the pseudocarpellary dorsals from 
posterior sepal and anterior petal strands are also characteristic of this group. Scaevola 
is also considered to be a member of this group in which the spur has been lost. The 
vascular pattern is still essentially the same except in the case of S. ramosissima, 
although the condensation of the vasculation is particularly noticeable. There is no 
separate derivation of petal laterals and staminal strands from the receptacular stele 
as in the other members of the Goodenia group, and the bundles only separate at the 
ovary dome. Thus Scaevola, in general, exhibits more complete and presumably more 
efiicient fusion than the rest of the group. Leschenaultia tends to stand even further 
apart from the rest of the family than Brunonia. The vasculation is, in many instances, 
quite different from any other member, as has been indicated above. The derivation of 
the pseudocarpellary dorsals and the presence of three large, closed spurs may indicate 
a relationship with the next group, but in other respects, notably the hollow style and 
the structure of the indusium (Carolin, unpub. ), it is even more divergent. 

Anthotium and Dampier-a have much in common and are quite different from the 
other groups. The derivation of the pseudocarpellary dorsals and the presence of several 
spurs distinguish them from the Goodenia group, and the ribs on the pseudo-ovary 
wall and the placental supply separate them from Leschenaultia, in addition to the 
more orthodox characters. Anthotiuin shows considerable fusion between whorls, but 
the vasculations of the different organs tend to remain more or less separate, at least 
in the upper part of the pseudo-ovary (Fig. 45, 37). In Dampiera the ovular number 
and the size of spurs are reduced. In addition, the condensation of the vascular supply 
is more evident than in Anthotium (Fig. 45, 38), and the ovary becomes "unilocular" 
by reduction of the basal "bilocular" zone and solidiflcation of one "loculus". 

It is evident that the scheme of affinities put forward by Krause (1912) has to be 
considerably modified in the light of this investigation. 

Extrapolating these characters, an ancestral form can be suggested (Fig. 45). It 
was presumably pentamerous except for the gynoecium which consisted of four united 
carpels; the petals were united into a tube as in Brunonia and the sepals were free. 
From this, Brunonia has developed by the reduction of the ovary (possibly in the same 
manner as the other genera, although equally possibly by simple abortion of two 



254 



FLOEAL STRUCTURE AND ANATOMY IN THE FAMILY GOODENIACEAE, 



carpels). The other genera have developed after a slit appeared in the corolla and 
further development has been along the usual lines of ovular reduction, gradual increase 
in union between floral parts and condensation of their vascular systems. On these 



diaspasis 
[k^^c'.s^aSg'.'"] 




BRUNONIA 



VERRAUXIA 



[kV^//^V^^] j 



Text-figure 45. Phylogeny of floral types in the Goodeniaceae. 
• Sepal Traces ; [j Petal Traces ; X Stamen Traces ; Q Carpel Traces. 

results it is impossible to suggest any relationships between this family and any 
other modern groups. The pollinating mechanism would appear to be the connecting 
link between Goodeniaceae and Campanulaceae, and it is hoped to make this the 
subject of a separate paper. 



BY R. C. CAKOLIN. 255 

If these phylogenetic schemes bear examination from other quarters, the family, in 
the wide sense, may be divided into the following subfamilial groups: (i) Goodenia 
group: Velleia, Symphyobasis, Goodenia, Calogyne, Selliera, Verreauxia, Diaspasii, 
Scaevola (Pentaptilon and Catosperma) ; (ii) Leschenatjltia group: Leschenaultia; 
(in) Dampiera group: Anthotium, Dampiera; (iv) Brunonia group: Brunonia. Cyto- 
logical evidence (Martin, Peacock, Carolin, all unpub.; Jackson, 1958) and other morpho- 
logical observations all tend to support this conclusion. 

These results also suggest that certain generic limits may need revision. Calogyne 
berardiana, as it differs so little from Goodenias and as its style can no longer be 
considered to be characteristic, should probably be returned to Goodenia. Scaevola 
fasciculata and S. helmsii should probably be referred to Verreauxia. 

Acknowledgements. 
The collection of much of the material used in this investigation was made possible 
by a Research Grant from the University of Sydney. The Directors of the Royal Botanic 
Gardens, Sydney, and National Botanic Gardens, Melbourne, have kindly donated 
material, and Prof. C. L. Wilson of Dartmouth College collected some material in 
Western Australia which has been used. 

Bibliography. 
Brough, p., 1927. — Studies in the Goodeniaceae, i. The life History of Dampiera stricta R.Br. 

Proc. Linn. Soc. N.S.W., 52: 471. 
CoLOzzA, A., 1907 and 1908. — Studio Anatomico sulle Goodeniaceae. Nuov. Gior. Bot. Ital., N.S., 

14: 304 ; 15 : 5. 
Douglas, G., 1944. — The Inferior Ovary. Bot. Rev., 10: 125. 

, 1957. — The Inferior Ovary. II. Bot. Rev., 23 : 1. 

Eambs, a., and MACDANiEfLS, L. H., 1947. — An Introduction to Plant Anatomy, ed. 2, New York. 

Hall, B. A., 1956. — Problems and Methods in Floral Anatomy. Phytomorph., 6 : 123. 

Jackson, W. D., 1958. — Chromosome numbers in Tasmanian Goodeniaceae and Brunoniaceae. 

Papers and Proc. Roy. Soc. Tas., 92 : 161. 
Krausb, K., 1912. — Das Pflanzenreicli. Goodeniaceae and Brunoniaceae. Berlin. 
PrrRl, v., 1951. — The role of floral anatomy in the solution of morphological problems. Bot. 

Rev., 17 : 471. 

, 1952. — Floral Anatomy and Inferior Ovary. Pliytom,orph., 2 : 122. 

Saunders, E., 1939. — Floral Anatomy, vol. 2, Cambridge. 



256 



IRON DEFICIENCY IN EUCALYPTUS DIVES SCHAUER. 

By W. D. Andrew and D. J. David, Division of Plant Industry, C.S.I.R.O., Canberra. 

(Communicated iy Professor L. D. Pry or.) 

[Read 29th July. 1959.] 



Synopsis. 

A chlorosis In Eucalyptus dives Schauer growing at Ginninderra, A.C.T., was observed 
during the spring of 1956 when rainfall was excessive and soils became waterlogged. It 
disappeared when the soils dried out and did not reappear in 1957 or 1958 when rainfall •w&.s 
below normal. 

Treating affected' leaves with iron compounds reduced the intensity of the chlorosis. 

Spectrographic analyses of chlorotic foliage in 1956 and healthy foliage in both 1956 and 
1957 showed that Ihe iron content of various portions taken from chlorotic branches was 
generally lower than that of corresponding parts of healthy branches from both 1956 and 
1957 samples. 



Inteoduction. 

Eucalyptus dives Schauer is widespread in Australia, grows readily on poor soils 
and dry ridges (Ewart, 1930) and is not ordinarily found in poorly drained situations 
(C. W. B. Moore, private communication). 

Pryor (1956) reported that this species was one of a relatively small group 
(Renantherae) of Eucalypts that readily developed chlorosis if grown in steam 
sterilized soil but was healthy in the same soil unsterilized. He concluded that this 
chlorosis was due to the absence of mycorrhiza on the roots following sterilization of 
the soil. 

During the late winter of 1956, chlorotic foliage was observed on a number of 
adult trees of E. dives growing at Ginninderra in the Australian Capital Territory. 
The affected leaves were the younger ones occurring mostly on the northern and north- 
western sides of the trees. 

The rainfall at Ginninderra in 1956 was nearly twice the normal amount of 
approximately 24" p. a. and soils were waterlogged during the winter. Many species of 
trees in the Australian Capital Territory suffered from excessive wetness and some died. 

Experimental Procedure. 

During the spring of 1956, chlorotic leaves on an affected tree were matched for 
colour, labelled, and six at random were dipped into a solution of ferric potassium 
ethylene-diamine tetra acetate (0-2 ppm iron) containing a small amount of Tween-20 
spreader. 

A number of chlorotic leaves were detached, six were partly immersed in distilled 
water and six in a solution of ferrous sulphate (0-04 ppm iron) in petrie dishes in the 
laboratory. 

Six branches bearing normal and six having chlorotic leaves were detached from 
trees and separated into current seasonal growth and previous seasonal growth as 
indicated by the colour and development of the main stem. The older portion was 
further subdivided into leaves and peduncles. 

In 1957 no chlorotic leaves were evident and branches were taken at random 
dijring the spring when the trees were at the same growth stage as those sampled in 
1956. These were separated into bulk fractions as in the previous year. 

The samples were analysed spectrographically, standards for the analyses being 
prepared by adding known amounts of the elements under analysis to portions of a 
composite sample of dry matter. Sulphated ash of the samples and standards were 

Proceedings of the Linnean Society of New South Wales, 1959, Vol. Ixxxiv, Part 2. 



BY W. D. ANDREW AND D. J. DAVID. 



257 



packed into cavities in graphite electrodes and arced at 15 amps d.c, the lower ash- 
fllled electrode being the anode. The spectrograph used was a Hilger Large Quartz 
instrument set at the wave-length range 2450-3500 A, the plates were Ilford Thin Film 
Half-tone and development was effected using ID13 developer. Analysis line intensities 
were computed from line densities measured on a Hilger non-recording microphotometer. 
The coefficient of variation associated with this method of analysis is about ± 10% for 
a single determination. The estimates of concentration of the elements sought were 
made on the means of triplicate spectrographic determinations on each ash sample. 

Results. 
Four weeks after treatment with Fe-K-EDTA the chlorotic leaves on the trees 
had developed green patches of colour in places where injury to the cuticle had occurred 
from various causes. The untreated leaves with similar injury to the cuticle remained 
chlorotic. 

Table i. 

Content of Iron, Manganese and the Iron I Manganese Ratio in the Dry Matter of Healthy and Chlorotic 

Foliage of Eucalyptus dives Schauer from Ginninderra, A.C.T. 



Sample. 


Fe. 


Mn. 


Fe/Mn 




(P.p.m.) 


(P.p.m.) 


Ratio. 


A. Current seasonal growth. 








1956 (wet year) : 








/ Healthy foliage 
\ Chlorotic foliage 


95 


620 


0156 


45 


870 


053 


1957 (dry year) : 








Healthy foliage 


70 


449 


156 


jB. Previous seasonal growth. 








1956 (wet year) : 








r Healthy leaves 
\ Chlorotic leaves 


85 


627 


135 


40 


404 


0-099 


J Healthy peduncles 
\ Chlorotic peduncles 


85 


449 


0-189 


62 


888 


0-070 


1957 (dry year) : 








Healthy leaves 


155 


745 


0-208 


Healthy peduncles 


175 


790 


0-221 



In the laboratory, the leaves immersed in the solution of ferrous sulphate became 
greener after about 10 days, the mid-ribs darkened and numerous dark spots developed 
near their apices. Those immersed in water did not change colour or develop spots. 

Table 1 gives the results of spectrographic determination of the iron content of 
various portions taken from chlorotic and healthy branches. It will be noted that it 
was lower in chlorotic samples taken in 1956 than in healthy samples taken in both 
1956 and 1957. 

As it is known that reduced uptake of iron under waterlogged conditions is often 
associated with excessive uptake of manganese (Twyman, 1946), the manganese con- 
centrations were also determined and these figures are included in Table 1. 

Discussion. 

Since the analyses showed that the unhealthy parts were lower in iron than the 
corresponding healthy ones, and the treatment of the chlorotic foliage with iron com.- 
pounds caused greening to occur, it is evident that the trees were suffering from an 
iron chlorosis in 1956 (Karschon, 1956). 

As Leeper (1935) has shown that the availability of soil manganese can be greatly 
increased during a wet season it is to be expected that in 1956 trees in the Australian 
Capital Territory had a greater opportunity for taking up an excessive amount of 
manganese than usual. 

Smith and Specht (1953) have shown that a reduction in the Fe/Mn ratio is 
characteristic of iron chlorosis induced by an excessive uptake of manganese. The 



258 IRON DEFICIENCY IN ElfCALYPTUS DIVES. 

ratios given in Table 1 are not inconsistent with this, but it should be noted that the 
manganese concentration was not in every case higher in chlorotic than in healthy 
foliage. 

McCool (1935) has shown that manganese induced iron chlorosis is more severe 
in high light intensity and Wiederspahn (1957) showed that these symptoms in apples 
developed first on the better lighted sides of the trees. Since the iron chlorosis observed 
in E. dives at Ginninderra occurred on the northern and north-western sides of the 
trees, it, also, was probably dependent on incident light intensity. 

References. 

EwART, A. J., 1930. — Flora of Victoria. Melb. Uni. Press. 

Kauschon^ R., 1956. — Iron chlorosis of Eucalyptus camaldulensis Dehn., and its incidence in 
relation to soil conditions. Ilanoth (3). Publ. Dept. Forests, Israel. 

Leepbr, G. W., 1935. — Soils and manganese deficiency. J. Aust. Inst. Agric. Sci., 1 : 161-163. 

McCooL, M. M., 1935. — Effect of light intensity on the manganese content of plants. Boyce 
Thompson Inst. Contr., 7 : 427-437. 

Peyor, L. D., 1956. — Chlorosis and lack of vigor in seedlings of Renantherous species of 
Eucalyptus caused by lack of mycorrhiza. Proc. Linn. Soc. N.S.W., 81: 91-96. 

Smith, P. F., and Specht, A. W., 1953. — Heavy metal nutrition and iron chlorosis of citrus 
seedlings. Plant Physiol, 28 : 371-382. 

Tv/YMAN, E. S., 1946. — The iron manganese balance and its effect on the growth and develop- 
ment of plants. New Phytol., 45 : 18-24. 

Wiederspahn, F. E., 1957. — Controlled manganese deficiency of apple. Proc. Amer. Soc. Hort. 
Sci., 69 : 17-20. 



259 



POLLEN TETRAD SEGREGATION IN A8TR0L0MA PINIFOLIUM AND IN 
ACROTRICHE FASCICULIFLORA. 
By S. Smith-White, Botany Department, University of Sydney. 
(Four Text-figures.) 

[Read 26th August, 1959.] 



Synopsis. 
An analysis is made of tetrad segregation in the -two species on the bases of three 
hypotheses — independence, the operation of cytoplasmic competition, and a relationship to the 
two divisions of meiosis. It is concluded that both cytoplasmic and meiotic conditions may 
contribute to the behaviour. There is also a significant contribution by environmental factors, 
and a precise partition of the effects of the environment, the cytoplasm, and the nucleus, is 
not yet possible. 



Inteodtjction. 

Pollen tetrad segregation, in which some or all of the pollen grains in a tetrad are 
aborted, has been recorded in twelve species of the Styphelieae (Smith-White, 1959). 
From a consideration of the patterns of pollen development found in other species of 
the tribe, it has been inferred that this tetrad segregation must have originated on 
several occasions and that these origins have been dependent upon a prior evolutionary 
history of monad-type pollen development and cytoplasmic polarity in pollen mother 
cells. In fact, it is inferred that within the Styphelieae there has been preadaptation 
tG the establishment of pollen tetrad segregation. 

It is reasonable to expect that the characteristics of such behaviour might reflect 
its causation. An analysis of the frequencies of the five possible kinds of pollen 
tetrad — nullads, monads, dyads, triads and full tetrads — may contribute to an under- 
standing of the mechanisms involved. Unfortunately, the tetrahedral arrangement of 
the pollen grains in the tetrads, and the distortion due to the differential growth of 
good and aborting grains, prevent the recognition of the planes of the first and second 
meiotic divisions. Analysis must be made on the basis of unordered tetrads. 

In this paper an analysis has been attempted of data for two species, Astroloma 
pinifolium Benth. and Acrotriche fasciculiflora Benth. 

Ast. pinifoliuni occurs in coastal districts of eastern Australia, extending from 
southern Queensland to eastern Tasmania. In this distribution it shows several major 
disjunctions, and it usually occurs in small isolated local populations in each region. 
Still more isolated populations are found in the Warialda district of north-west New 
South Wales and in the Grampians of south-western Victoria. The pattern of distribu- 
tion suggests that the species is old and relict. Material studied in this paper was 
obtained from the east coastal region, from Sydney to Evans Head. Data from 
Grampians plants are presented for comparison and contrast. 

Acr. fasciculiflora is found only in south-eastern South Australia, its range 
extending from the Adelaide Hills and Mt. Lofty Ranges to Kangaroo Island. The 
material studied was collected in Long Gully, Belair National Park. 

The Data. 

Observed frequencies of tetrad types for a representative selection of plants of the 
two species are given in Tables 1 and 2. Table 3 summarizes the analysis according 
to two hypotheses for 57 plants of Ast. pinifoliutn. Tables 4 and 5 present data and 
analysis for within-plant variation for two selected plants of Ast. pinifolium. 

The data given for each plant in Tables 1 and 2 are composite in that they were 
obtained from two or more fiowers and 10 or more anthers. There is significant 
variation between flowers and between anthers within plants, which is illustrated by 
the data given in Tables 4 and 5, and by the graph in Text-figure 3. 

Proceedings of the Linnean Society of New South Wales, 1959, Vol. Ixxxiv, Part 2. 



260 



POLLEN TETEAD SEGREGATION IN ASTKOLOMA AND ACROTRICHE, 



Table 1. 

Astroloma pinifolium. 

Tetrad Segregation Data. 







Tetrad Type Proportions. 










Plant and 






% X 100. 






N. 


(p + q) or 


x'- 


Hypothesis. 














(x+y+z). 


















To. 


Ti. 


Ta- 


r,. 


Tj. 








B.57/8 


7673 


1790 


484 


52 


1 


9595 






I .. 




73S4 


2326 


273 


15 







0-927 + 0-073 


378-2 


Ill . . 




7658 


1808 


485 


44 


5 




0-8751+0-1033+0-0216 


0-49 


LP57/4 




7183 


2347 


415 


54 





4651 






I .. 




7056 


2571 


351 


21 







0-9165 + 0-0835 


38-1 


Ill . . 




7174 


2338 


444 


41 


2 




0-847+0-138+0-015 


2-2 


W54/6 




6011 


3151 


750 


83 


5 


5878 






I .. 




5916 


3319 


698 


65 


2 




0-877 + 0-123 


12-3 


Ill . . 




6006 


3147 


753 


89 


5 




0-775+0-203+0-022 


0-17 


B,57/5 




4982 


3538 


1262 


208 


9 


4468 






I .. 




4746 


3888 


1195 


163 


8 




0-83+0-17 


25-4 


Ill . . 




4970 


3596 


1214 


204 


16 




0-705+0-255+0-040 


2-7 


TG57/4 




4065 


4182 


1488 


243 


22 


8785 






I . . 




4116 


4091 


1524 


253 


16 




0-801+0-199 


5-2 


Ill . . 




4116 


4091 


1524 


253 


16 




0-642+0-318+0-040 


5-2 


LP53/3 




3998 


3728 


1866 


374 


34 


8022 






I .. 




3797 


4160 


1709 


312 


21 




0-785 + 0-215 


63-1 


Ill . . 




3969 


3780 


1782 


420 


49 




0-63 + 0-30+007 


n-5 


TG57/2 




31.57 


4204 


2116 


482 


41 


9595 






I .. 




3145 


4219 


2122 


472 


40 




0-7489+0-2511 


0-34 


Ill . . 




8165 


4204 


2114 


477 


41 




0-5626+0-3736+0-0638 


0-07 


055/2 




2914 


4313 


2301 


451 


21 


6088 






I ,. 




3015 


4215 


2210 


515 


45 




0-741+0-259 


18-3 


Ill . . 




2884 


' 4435 


2243 


413 


25 




0-537+0-413+0050 


5-5 


R57/6 




2874 


4138 


2321 


603 


63 


15284 






I .. 




2820 


4199 


2345 


582 


54 




0-7287+0-2713 


6-8 


Ill . . 




2862 


4119 


2338 


616 


64 




0-535+0-385+0-080 


0-8 


W54/5 




2659 


4332 


2504 


475 


30 


3603 






I .. 




2809 


4198 


2353 


.586 


55 




0-728+0-272 


19-5 


Ill . . 




2601 


4386 


2461 


516 


36 




0-51+0-43+-0-06 


2-5 


TG57/I 




2743 


3983 


2659 


580 


35 


9254 






I .. 




2695 


4182 


2433 


629 


61 




0-7205+0-2795 


42-9 


Ill . . 




2695 


4182 


2433 


629 


61 




0-519+0-403+0-078 


42-9 


W54/2 




2459 


4354 


2665 


501 


21 


3753 






I .. 




2658 


4175 


2460 


644 


63 




0-718+0-282 


37-2 


Ill . . 




2401 


4410 


2613 


540 


36 




0-49+0-45+0-06 


3-1 


W54/1 




2559 


4276 


2501 


625 


39 


3314 






I .. 




2643 


4173 


2470 


650 


64 




0-717+0-283 


5-4 


Ill . . 




2500 


4300 


2549 


602 


49 




0-50 + 0-43+0-07 


1-7 


TG57/5 




2970 


3472 


2635 


833 


90 


8923 






I .. 




2541 


4152 


2544 


693 


71 




0-710+0-290 


197-0 


Ill . . 




3025 


3740 


2366 


748 


121 




0-55+0-34 + 0-11 


61-5 


055/1 




2005 


3980 


3027 


906 


81 


1721 






I .. 




2052 


3987 


3906 


941 


114 




0-673+0-327 


3-1 


Ill . . 




1980 


4049 


2960 


910 


100 




0-445+0-455+0-100 


11 


A53/1 




2079 


3513 


3495 


857 


56 


11322 






I .. 




1991 


3959 


2951 


978 


121 




0-668+0-332 


230-0 


Ill . . 




1998 


3996 


2946 


948 


112 




0-447+0-447+0-106 


226-4 


W54/9 




1619 


3972 


3277 


1054 


78 


3463 






I . . 




1785 


3845 


3105 


1141 


150 




0-650 +0-350 


23-1 


Ill . . 




1600 


4016 


3304 


984 


96 




0-400 + 0-502 + 0-098 


3-1 


057/6 




1078 


4260 


3988 


635 


39 


9001 






I .. 




1709 


3796 


3160 


1170 


162 




0-643+0-357 


716-1 


Ill . . 




1156 


4114 


4029 


665 


30 




0-304+0-605+0-055 


12-6 


057/1 




1199 


3712 


3649 


1286 


154 


9424 






I .. 




1419 


3572 


3372 


1414 


223 




0-614+0-386 


38-2 


Ill . . 




1225 


3710 


3649 


1272 


144 




0-35 + 0-53+0-12 


1-2 



Localities : A, AudJey. R, LP, La Perouse and Malabar. O, Oatley. E, Evans Head. TG, Tea Gardens. 
W, Warrah. Gr, Grampians. 



BY S. SMITH-WHITE. 



261 



Table 1. — Continued. 

Astrnloma pinifolium. — Continued. 

Tetrad Segregation Data. — Continued. 







Tetrad 


Type Proportions. 










Plant and 






% X 100. 






N. 


(p + q) or 


X • 


Hypothesis. 














(x+y + z). 


















To. 


Ti. 


r.. 


r^. 


Ti- 








E51/2 


454 


3128 


4876 


1424 


ns 


1608 






I .. 


957 


3086 


3644 


1913 


377 




0-5594 +0-4406 


136-4 


Ill . . 


484 


2948 


4968 


1474 


121 




0-22+0-67+0-11 


2-55 


057/5 


940 


2464 


3698 


2351 


547 


6466 






I .. 


742 


2719 


3736 


2280 


522 




0-522+0-478 


52 1 


Ill . . 


841 


2668 


3566 


2300 


625 




0-29+0-46+0-25 


26-5 


055/8 


3169 


4260 


2191 


351 


29 


2392 






I .. 


3249 


4218 


2053 


444 


36 




0-755+0-245 


13-0 


Ill . . 


3136 


4368 


2081 


390 


25 




0-56 + 0-39+0-05 


3-1 


R57/11 


36.30 


3918 


1974 


434 


43 


9098 






I .. 


3452 


4206 


2107 


390 


36 




0-7665+0-2335 


39-5 


Ill . . 


3600 


3960 


1929 


462 


49 




0-60+0-35+0-07 


3-8 


GR58/1007 . . 


3028 


5946 


1026 








2758 


q = 0-200 




Gr58/1005 


1229 


7149 


1614 


4 


4 


2571 


q = 0-260 




Gr58/1008 


660 


1657 


7579 


33 


71 


2408 


q = 0-430 




Gr58/1001 


431 


2082 


7416 


5 


66 


1974 


q = 0-4.30 





Localities : A, Audley. K,, LP, La Perouse and Malabar. 0, Oatley, E, Evans Head. TG, Tea Gardens. 

W, Warrali. Gr, Grampians. 

Eor liypothesis I (independence) expected frequencies are given in italics wliere they are below observed 
frequencies. 

In the collection of the data, the greatest possible care has been taken to avoid 
biologically meaningless sources of error. In slide preparation, anthers were dissected 
in a drop of stain-mountant on the slide under binocular magnification, to remove at 
least 90% of the pollen. The drop of mountant used was limited so that no exudation 
occurred at the edges of the cover slips, which might lead to a loss of the small nullads. 
Finally, within-slide sampling error has been avoided in most cases by scanning the 
whole area of each slide. In the data given for the Grampians plants of Ast. j)inifolium 
a sectorial sampling method was used. 

The Analysis. 
Three different hypotheses have been examined. These are: (I) The survival or 
death of any microspore is independent of the fates of its three sister microspores iu 
the same tetrad. (11) There is competition for survival between the four microspores 
in each tetrad. (Ill) The fates of the four microspores in each tetrad are related to the 
meiotic divisions. 

I. The hypothesis of independence. 
With independence, the observed frequencies of the five classes of tetrad should 
conform to the terins of the binomial (p + q)*, where q is the prospect that any micro- 
spore will survive, and p (=l-q) is its prospect of abortion. 

Writing ao....ai for the observed frequencies of the five tetrad types, their 

summation being N, the total sample, and r, r, for the corresponding proportions, 

summating to unity, the observed value of q is 

4aj + Sag + 2a2 + a 



4N 
Using values of q obtained directly from the data, expected frequencies eo....ei, 
based on the hypothesis of independence, have been calculated. Comparison of the 
observations with the expected frequencies (Hypothesis I) are presented in Tables 1, 
2 and 3. Only one-quarter of the plants of eastern Ast. pinifolium show an acceptable 
fit to the hypothesis. The distributions of tetrad-type frequencies in the Grampians 

H 



262 



POIXEN TICTKAl) SEGREGATION IN ASTROLOMA AND ACROTRICHE. 



plants are remote from the binomial form, and in Acrotriche fasciculiflora. all tliirteen 
plants examined show an unsatisfactory fit. 

The hypothesis of Independence does not provide a generally satisfactory explana- 
tion of the segregation behaviour in either species. In the eastern A. pinifolium. 
however, it seems probable that the causes of tetrad segregation are occasionally 
capable of simulating independence. 



Table 2. 

Acrotriche fasciculiflora. 

Tetrad Segregation Data. 







Tetrad 


Type Proportions. 










Plant and 






% X 100. 






N. 


Parameters 


x^. 


Hypothesis. 














(p+'.i) or (x+y + z). 


















To- 


ri. 


r». 


rj. 


r,. 








.54/11 . . 


9603 


332 


62 


3 





3378 






I .. 


9544 


448 


8 










0-9884+00116 


213-8 


Ill . . 


9604 


333 


62 


1 







0-980 +0-017 +0-003 


0-03 


54/12 . . 


9012 


739 


244 


5 





4008 






I .. 


8813 


1132 


54 


1 







0-9689+0-0311 


318-9 


Ill . . 


9025 


722 


242 


9 







0-950 + 0-038+0-012 


0-05 


54/8 .. 


1824 


4368 


3231 


561 


16 


1891 






I .. 


2214 


4055 


2784 


850 


97 




0-686+0-314 


59-8 


Ill . . 


1806 


4420 


3172 


572 


30 




0-425 + 0-520 + 0-055 


1-68 


54/9 . . 


1331 


4429 


3775 


459 


6 


1778 






I ., 


1967 


3947 


2969 


993 


124 




0-666+0-334 


158 - 1 


Ill . . 


1369 


4403 


3799 


417 


12 




0-370+0-595+0-035 


0-76 


54/7 . . 


1010 


4657 


3872 


461 





2278 






I .. 


1841 


3878 


3064 


1076 


142 




0-655+0-345 


276-4 


Ill . . 


1156 


4250 


4144 


438 


12 




0-34+0-625+0-035 


17-2 


54/1 . . 


1349 


3678 


4074 


888 


11 


2602 






I .. 


1646 


3753 


3208 


1212 


174 




0-637+0-363 


120 


Ill . . 


1122 


4087 


4090 


671 


30 




0-335+0-610 + 0-0.55 


43-8 


54/3 . . 


974 


4273 


3980 


773 





1294 






I ., 


1636 


3746 


3216 


1227 


176 




0-636+0-364 


104-2 


in . . 


1069 


4009 


4150 


736 


36 




0-327+0-613 + 0060 


4-2 


54/4 . . 


833 


4395 


4155 


616 





1752 






I .. 


1638 


3746 


3216 


1227 


176 




0-636+0-364 


214 - 9 


Ill . . 


973 


4012 


4415 


579 


20 




0-312+0-643+0-045 


12-7 


54/2 . . 


822 


4183 


4300 


655 


40 


2979 






I .. 


1546 


3678 


3282 


1302 


196 




0-627 + 0-373 


348-1 


Ill . . 


961 


3968 


4406 


640 


25 




0-31+0-64+0-05 


13-1 


54/5 . . 


784 


4067 


4579 


606 





2326 






I .. 


1516 


3654 


3303 


]327 


200 




0-624+0-376 


345-3 


Ill . , 


824 


3846 


4736 


577 


18 




0-287 + 0-670+0-043 


5-9 


54/10 . . 


677 


3862 


4602 


854 


5 


2084 






I .. 


1375 


3533 


3402 


1456 


234 




0-609 + 0-391 


267-2 


Ill . . 


784 


3752 


4769 


670 


25 




0-28+0-67+0-05 


12-6 


54/13 . . 


600 


4122 


4277 


966 


35 


1999 






I .. 


1349 


3498 


3403 


1471 


239 




0-607+0-393 


218-7 


Ill . . 


812 


3733 


4732 


786 


36 




0-285+0-655 + 0-060 


32-7 


54'6 . . 


409 


3232 


5274 


1067 


18 


1640 






I .. 


1086 


3223 


3587 


1775 


329 




0-574+0-426 


278-1 


Ill . . 


484 


3168 


5448 


864 


36 




0-22 + 0-72+0-00 


10-8 



Contingencies : 

Total, x^'^'SS-l (exdiidiruj 54111 and 54112). 

Between plants, x^ = 542-5. 

Witliin plants, x2 = 255-6. 
Wittiin plant data not presented in table. 
All sets of data are from counts of two tlowers. 

(The full tetrad classes, being generally very small, have been grouped with the triad classes for the calculation of 
the contingency x^ leaving a total of 22 degrees of freedom.) 

II. The hypothesis of competition. 
There are many reasons for the inference that cytoplasmic gradients may constitute 
the primary system of control in pollen development in the Styphelieae (Smith-White, 
he), It is possible that the frequent occurrence of tetrad segregation in the tribe is 



BY S. SMITH-WHITE. 



263 



dependent upon cytoplasmic differentiation. Such a system would necessarily involve 
competition between sister microspores for cytoplasmic necessities. The occurrence of 
a competition effect, however, would not exclude a nuclear or even a chromosomal 
mechanism. 

Competition between sister microspores would cause deviations from the theoretical 
frequencies calculated on the basis of independence. Such competition may be either 
positive or negative. 

Negative competition or assistance would occur if, when some microspores survive, 
others in the same tetrad have an increased prospect of survival. A similar situation 
would exist if the death of a microspore increased the prospect of death of other 

Table 3. 
Maximum Likelihood Approximations to Independence and to the Trinomial Square Form. 
Bast Coa-st Populations of Astroloma pini/olium. 



Hypothesis. 


Population. 


Number of Plants showing x' 
Probabilities 


Values Equivalent to 
of 


Total 
Plants. 




<001 


0-01-0-05 


05-0 -10 


010-0-50 


0-50-0-90 


0-90-0 -95 


0-95-1 


I. Independence 
(p + q)' 


Oatley 

Warrah 

Evans Head 

Malabar 

Tea Gardens . . 


13 

6 

11 

2 


3 


1 


3 
1 


1 
2 


2 


2 
1 


21 
8 
6 

18 

4 




Totals 


39 


3 


1 


6 


3 


2 


3 


57 


III. Trinomial 
square 

(x+y + z)'' 


Oatley 

Warrah 

Evans Head 

Malabar 

Tea Gardens . . 


3 

1 
2 


3 
3 , 


— 


i 
2 
3 
3 
1 


6 
5 
2 
6 


2 
1 
1 
3 


3 

2 
1 


21 
8 
6 

18 
4 




Totals 


6 


6 





13 


19 


7 


6 


57 



members of the tetrad. Just this effect has been observed in an unusual strain of 
PeUmia studied by Levan (1942). It leads to an excess of nuUads and full tetrads, and 
a deficiency of the other classes, compared with independence. Levan infers that dying 
pollen grains produce toxic degenerative substances which have adverse effects on the 
other members of the tetrad. 

Conversely, positive competition implies that where some microspores develop, 
others in the same tetrad are at a disadvantage — necessary substances may be in limited 
supply, and may have non-uniform distribution within the mother cell. 

Comparison of the data in Table 1 with binomial expectancy shows that in eastern 
Ast. pinifolium, in those plants where the data do not fit the hypothesis of independence, 
there is no consistent surplus or deficiency in the dyad and monad classes as compared 
with the binomial expectancy. There is no consistent indication of either positive or 
negative competition within tetrads. 

In the data for Ac?: fascicicliflora given in Table 2, however, all thirteen plants 
showed a marked excess of dyads over binomial expectancy, nine plants show an excess 
of monads, and all other classes in all plants, with the minor exception of nullads in 
plant 54/12, are deficient. The consistency of the thirteen sets of data suggests very 
strongly that competition between microspores, does operate, and further analysis is 
justified. 

If Qi and Pi, respectively, represent the prospects that a first microspore will survive 
or abort, and if q., qj, q^ and p„, Pj, p^ are the prospects for survival or abortion of a 
second microspore following the survival of a first, of a third following the survival of 
two, and of the fourth following the survival of the other three, the theoretical 



264 POLLEN TETEAD SEGREGATION IN ASTEOLOMA AND ACEOTEICHE, 

frequencies of the five tetrad categories will be: 
eo = Pi' 
ei = Cb ( Pa' + Pi'Pa + P1P2* + W ) 

62 = qiQb (Pi' + P2' + Pa' + P1P2 + P1P3 + P2P3) 

63 = qiq2q3 (Pi + P2 + Ps + PJ 
e.i = qiqoqsq*. 

The values q^, qo, qg and q^ can be related such that 
q.. = fiqi, q3 = fsQa, q* = Ufls 
and fo = afi, i, = caf„ = ca%. 
Dropping the subscript to fj, 

q2 = fqi, q3 = af^qi, q4:=ca'f'qi. 
In this formulation, there are four parameters, qj, f, a and c, and the last three 
measure or determine the increasing severity of competition between first and second, 
second and third, and third and fourth microspores. Since there are at most four 
degrees of freedom in any set of data, statistical testing of the formulation is not 
possible. However, values for the parameters have been calculated for eleven sets of 
data from Table 2 (excluding plants 54/11 and 54/12). The following five sets are 
representative : 

Plant q f a c 

54/8 -367 -841 -958 -133 

54/7 -436 -648 -650 -0077 

54/1 -394 -951 -562 -0133 

54/3 -441 -718 -794 -0000 

54/6 -550 -732 -582 -0588 

It will be noted that a is of the same order of size as f; it may be slightly larger, 
or substantially smaller. The parameter c is very small, but lower values of c are not 
necessarily associated with low values of a or of f. Competition acting against the 
third microspore is scarcely greater than };hat acting against the second, but that 
operating against a fourth microspore is very severe. 

The very sharp increase in competitive effect operating against the survival of all 
four microspores, which is indicated by this analysis, suggests that competition does 
not conform to a simple pattern. It is consistent, however, with the operation of a 
cytoplasmic gradient (Text-figure lA, B). If there is a gradient distribution of neces- 
sary cytoplasmic substances (stippling) and also variation in total amount, and if the 
plane of the first meiotic division tends to lie along this gradient, varying from position 
(a) to position (b) with a mean at (m), competitive effects consistent with those 
observed could result. It is perhaps significant that in Astroloma and in many other 
species of the Styphelieae the spindle of the first meiotic division is eccentric and 
laterally displaced (Text-figure IC) and may be across a cytoplasmic gradient. 

III. The hypothesis of meiotic dependence. 
■ This hypothesis assumes that the tetrad segregations are related to and dependent 
upon the conditions of meiosis. Martin and Peacock (1959, pp. 272-278) have examined 
tetrad segregation in several species of Leschenaultia (Goodeniaceae). They find 
that, consequent upon low chiasma frequency and terminal chiasma localization, 
there is an appreciable frequency of univalents at first metaphase in pollen mother 
cells; loss and misdivision of univalents is responsible for some, but probably not all, 
of the observed pollen grain death. 

In the Styphelieae it is not possible to recognize the plane of the first division in 
the mature pollen tetrads. This plane, however, may still be considered to divide the 
pollen mother cell (PMC) into two half-mother cells (i MCs). Each i MC would then 
be capable of yielding one of three possible results. — neither, one only, or both its 
daughter microspores might survive. The prospects of these three possible results will 
be designated by the symbols x, y and z respectively. 

With independence, there would arise the identities x = p"; y = 2pq; z = q^; and 
(x-fy + z)^:: (p + q)*. 



BY S. SMITH-WHITE. 



265 



Independence, however, might operate between, but not within, half-mother cells 
and the tetrad-type frequencies would then fit the expansion of the trinomial (x-t-y-fz)'. 
This relationship, v/hich will be referred to as the "trinomial square", is illustrated in 
Text-figure 2. Recognition of the plane of the first division, if possible, would permit 
the separation of the terms y^ and 2xy in the theoretical frequency of dyads. This 
hypothesis of independence between i MCs is hypothesis III in Tables 1, 2 and 3. 




Text-figure 1. — The effect of competition in a cytoplasmic gradient. 

The gradient is represented by stippling. Variation in total supply of the gradient 
substances is not indicated. The orientation of the division spindles varies, with a mode as in 

B. Orientation as in A would yield monads or triads, according to the quantitative supply of 
gradient substance. Orientation as in B would yield dVads rather than monads. Cf. text. In 

C, the spindle eccentricity characteristic of the family is illustrated. 



In order to test the data in Tables 1, 2 and 3 against the hypothesis, maximum 
likelihood estimates of x, y and z have been approximated, using a method of iteration 
described by Kempthorne (1957, pages 172-177), the iteration being repeated only until 
a satisfactory x" value was obtained, or until it became obvious that a satisfactory 
x' value was unobtainable. In many of the sets of data, the initial estimates, obtained 
from the equations 



,=v 



2ro+ri-fr2— 2Vro, x=-J-, z=l— x— y 



have given acceptable x~ values, and no further iteration has been carried out. 

The observed frequency distributions obtained for eastern Ast. pinifolium are 
surprisingly consistent with this third hypothesis (Tables 1, 3). By contrast, the 
tetrad-type distributions obtained for the Grampians population of the species are 
remote from the form, and it must be inferred that the direct causes of segregation 
are quite different. 

In Acr. fasciculiflora, four plants among thirteen give an excellent fit to the 
hypothesis, two show a barely acceptable fit, and seven- show a very bad fit. 



266 



POLLEN TETRAD SEGREGATIOiV IN ASTROLOMA AND ACROTRICHE, 



Within-Plant Variation. 
Significant variation occurs between flowers within plants in probably all plants 
of both species. Intraplant variation, however, generally seems to be of lesser degree 
than interplant variation, and this feature is clearly shown by Text-figure 3. In Table 2, 

THE TRINOMIAL SQUARE FORM 






xy 












Text-figure 2. — The trinomial square form. 

X, y and z are the prospects for the three possible results from a i MC. With independence, 
the whole tetrad result becomes (x + y + z)^. 

the contingency x' value for within-plant variation is only one-half that for between- 
plant variation, for equal degrees of freedom. 

In Tables 4 and 5, data from individual flowers and anthers are given for two 
plants of Ast. pinifolium and are sufficient to illustrate the characteristics of intra- 
plant variation. These two plants have been selected as representative of plants 
showing good and bad flt, respectively, to the trinomial square form. 

In both plants there is significant variation, even between the five anthers of a 
single fiower. In R57/6 (Table 4), the grand totals, fiower totals, and individual anther 
sets, with one exception, all conform to trinomial squares. In A53/1, the data for 
flower 1 show an acceptable flt, but the other flowers do not, and the fit of the grand- 
total distributions is very bad. 

The data in Table 4 show two features which deserve comment. The first is that 
x' values tend to be either very low, corresponding to probabilities above 0-5, or very 
high with probabilities below 0-01. The distribution of x^ values does not appear to be 
normal. The second feature is that, although the individual anther distributions all flt 
trinomial squares, they are significantly different, and yet the grand total distribution 
conforms to a "mean" trinomial square. It can be shown algebraically that the means 
of the individual terms of two different trinomial squares cannot be themselves the 
terms of a trinomial square. This contradiction in the data has not been resolved. 

Intraplant variation in tetrad segregation is generally at a significant level, and 
must mean that the behaviour is subject to environmental modification. Interplant 
variation, however, is generally greater in degree, and this fact suggests that there 
are both genetical and environmental components involved. The existence of the 



SY S. SMITH-WHITE. 



267 



environmental effect, however, makes the interpretation of interplant variation difficult 
and hazardous. 

The Meaning of the Trinomial Square Form. 

The consistency of the fit to the trinomial square hypothesis in plants of eastern 
Ast. pinifoliuvi must have significance in relation to the causes of tetrad segregation. 
This form has several implications which relate it to the conditions of meiosis (Text- 
figure 2). 









O 
a 



50 - 



40 



30 



20 



10 - 




I I I 



J I 1 I I I I 






J \ I I 1 I I I I 



~y]ro■<tun^or^■«3■CD 

,^ r^i^fnr^]^'^??r^r^ 

m LD m tf) in in i^ m m m in m m inminini^inmin^^'^ 
000$<cc;:250^B:a:O^Q:Q:§^Q:5a:QjQ: 



PLANT NUMBER 



Astroloni,a 



Text-figure 3. — Variation in pollen fertility within and between plants. 
pinifoUum. 

The values plotted are the q values for total microspore survival, each determination 
being from the five anthers of a flower. It is suggested that both genetic and environmental 
components affect pollen fertility. 



The values x, y and z represent the prospects that neither, one only, or both the 
microspores derived from a second division 1 MC shall survive. It is clear that to give 
the trinomial square form: 

(i) The events leading to microspore failure or survival are operative or effective 
during the second division of meiosis. 

(ii) The two I MCs of each PMC must have similar prospects. 

(iii) These prospects in the two J MCs must he independent. 

(iv) From (ii) and (iii) it follows that the first meiotic division must be strictly 
equational, both in chromosomal segregations or abnormal chromosome behaviour, and 



268 



POLLEN TETRAD SEGREGATION IN ASTROLOMA AND ACROTRICHE, 



in cytoplasmic differentiation. In particular, the cytoplasmic polarity present during 
the first division in Leucopogon junipermus, which is demonstrated by the non-random 
segregation of univalents (Smith-white, 1948), and which is generally involved in the 
monad-type of pollen development in the Styphelieae (Smith-White, 1955, 1959), must 
be absent, or must be delayed, at least, until after the beginning of the second divisions. 

Table 4. 

Variation in Tetrad Segregation Within and Between Flowers. 

Plant 1157/6. 



Flower. 


Anther. 


Tetrad Type Proportions. 
% X 100. 


N. 


q. 


(x+y + z). 


X^ 


P. 




To. 


r,. 


Ta. 


Ts. 


Tt. 




1 


1 

2 
3 
4 
5 


457 
290 
279 
287 
248 


412 
458 
435 
399 
403 


118 
219 
243 
241 
246 


13 
31 

41 
66 

87 




2 

2 

7 

16 


608 
576 
581 
544 
549 


0-172 
0-247 
0-263 
0-277 
0-306 


0-676+0-304+0-020 
0-538+0-426+0-036 
0-528+0-412+0-060 
0-54+0-37 + 0-09 
0-50+0-40+0-10 


0-15 
0-10 
0-60 
0-19 
2-21 


0-98 

0-99 

0-8 

0-98 

0-7 


Totals 


315 


421 


212 


47 


5 


2858 


0-252 


0-562+0-375+0-063 


1-16 


0-85 


2 


1 
2 
3 
4 
5 


339 
282 
272 
259 
256 


356 
439 
431 
436 
423 


246 
229 
221 
242 
246 


59 
43 
69 
48 
72 



7 
5 
15 
3 


675 
674 
605 
665 
601 


0-256 
0-264 
0-276 
0-281 
0-286 


0-52+0-40+0-08 
0-52+0-40+0-08 
0-52+0-41+0-07 
0-50+0-40+0-10 
0-51+0-42 + 0-07 


191 
6-7 
2-4 

10-9 
1-85 


0-001 

0-15 

0-5 

0-03 

0-75 


Totals 


283 


416 


237 


58 


6 


3220 


0-272 


0-532 + 0-391+0-077 


0-44 


0-97 


3 


1 
2 
3 
4 
5 


328 
295 
231 
222 
123 


407 
423 
419 
413 
434 


223 
231 
257 
274 
317 


38 
49 
85 
82 
113 


4 
2 
8 
9 
13 


445 
529 
611 
559 
602 


0-246 
0-260 
0-305 
0-311 
0-365 


0-573+0-355+0-072 

0-55+0-39+0-06 

0-481+0-436+0-083 

0-47+0-44+0-09 

0-38+0-50+0-02 


1-96 
1-11 
2-09 
0-11 
7-65 


0-7 

0-8 

0-6 

0-99 

0-07 


Totals 


233 


420 


262 


76 


8 


2746 


0-301 


0-483+0-435+0-082 


1-67 


0-7 


Grand totals 


278 


419 


237 


60 


6 


8824 


0-275 


0-. 530 +0-395 +0-075 


1-36 


0-85 



(Full tetrad classes have been grouped with triad classes.) 
Contingencies: Total, x'' = 346-3. 42 d.f. 

Between flowers x° = 71-3. Heterogenous. 

Contribution to total x" by anthers 1 — 1 and 3 — 5 = 251-2. 

The trinomial square form of frequency distribution does not necessarily imply 
but it does suggest that chromosomal or genetical segregations are involved. It is 
possible to devise model systems which are capable of giving this form. The conditions 
necessary are: 

(i) There must be a substantial degree of chiasma localization, such that definite 
chromosome segments are always distal to the first chiasma, and always proximal to 
a second chiasma if such is formed. These segments will always show second division 
segregation and, like the differential segments of an Oenothera system, will be excluded 
from crossing over, and so protected from breakdown (Text-figure 4). 

(ii) These segments, on different chromosomes, must carry all genes, or chromosome 
blocks, concerned in the tetrad segregation. 

(iii) There must be at least two gene or chromosome block systems. A/a, B/b, . . . 
and A'/a', B'/b', . . . which may be either alternate or duplicate to each other in action. 
Each system must include two or more complementary genes or blocks, and all must 
be independent of one another in segregation. 

The minimum system possible is A/a B/b A'/a' B'/b' which would give 

A B and A' B' systems duplicate, (-125 + -625 +-250)^ 

A B and A' B' alternate, (-250 + -625 + -125)1 



BY S. SMITH-WHITE. 



269 



A system A/a B/b C/c A'/a' B'/b' C'/c' would permit a wider range of trinomial 
squares. 

Such an hypothesis of two gene or chromosome block systems is permitted by the 
apparent secondary polyploid constitutions of Astroloma pinifolium (n = 7, x = 4) and of 
Acrotriche fasciculiflora (n = 9. Smith-White, 1955). However, the hypothesis would 







Region of i a ' Region of ] 
lst.*Chiasma ' ' 2nd. Chiasma ' 

Text-figure 4. — Diagram of a bivalent with chiasma localization. 

Segments outside the chiasma regions are isolated from crossing over. The region cCistai 
to the first chiasma region would show regular reduction at second anaphase. 



not seem to permit the degree of intraplant and environmental variation which has been 
described, unless it were associated with a very considerable effect of environment on 
chiasma frequency and chiasma localization. 

Chromosome fragmentation is known to occur as a characteristic feature of meiosis 
in pollen mother cells of eastern Ast. pinifolium, and will be described in another paper. 
This abnormal meiotic behaviour may be associated with a system of chromosome 



TABLE 5. 

Variation in Tetrad Segregation Within and Between Flowers. 

Astroloma pinifolium. Plant A53/1. 



Flower. 


Anther. 


Tetrad-type Proportions. 


N. 


q- 


Parameters 

(xyx). 


X'. 


P. 


To. 


Tl- 


Ts. 


Tj- 


r,. 


1 


1 
2 
3 
4 
5 


332 
247 
254 
266 
231 


395 
468 
448 
442 
453 


239 
254 
270 
266 
265 


33 
31 

28 
26 
49 






2 


461 
460 
496 
421 
433 


0-242 
0-267 
0-268 
0-269 
0-282 


0-54+0-41+0-05 
0-50+0-46+004 
0-51+0-45+004 
0-51+0-45+0-04 
0-50+0-45+0-05 


6-1 
0-6 
2-9 
2-8 
1-0 


0-10 
0-85 
0-40 
0-40 
0-80 


Totals 


267 


441 


259 


33 





2271 


0-265 


0-51+0-45+0-04 


5-7 


0-15 


2 


1 
2 
3 
4 
5 


202 
185 
178 
134 
120 


326 
322 
301 
308 

277 


385 
368 
363 
429 
450 


82 
118 
148 
124 
143 


6 
6 

11 
5 

10 


491 
475 
535 
396 
476 


0-341 
0-360 
0-379 
0-.390 
0-412 


0-405+0-515+0-080 

0-39+0-52+0-09 

0-39+0-52+0-09 

0-32+0-58+0-10 

0-31+0-57+0-12 


18-5 
14-5 
17-7 
9-0 
14-2 


0-001 

0-004 

0-001 

001 

0-01 


Totals 


165 


307 


397 


123 


8 


2373 


0-376 


0-343 +0-550 +0-107 


82-9 


0-000 


3 


1 
2 
3 
4 
5 


140 
151 
135 
119 
102 


315 
336 
336 
330 
303 


381 
401 
405 
385 
458 


149 
101 
172 
152 
131 


15 
11 
10 
15 

7 


549 
456 
398 
455 
413 


0-363 
0-371 
0-386 
0-404 
0-409 


0-35+0-52 + 0-13 
0-36 + 0-53+0-11 
0-36+0-53+0-11 
0-34+0-55+0-11 
0-32+0-58+0-10 


7-6 

7-34 

4-78 

6-4 

9-1 


0-10 
0-10 
0-30 
0-15 
0-01 


Totals 


129 


323 


403 


133 


12 


2271 


0-393 


0-34+0-55+0-11 


26-4 


0-001 


Grand totals 


187 


356 


354 


97 


7 


6915 


0-345 


0-38+0-53 + 0-09 


130-1 


0-000 



Contingency ; 



Total x' 
Between flowers 
Within flowers 



533-4. 

439-3. 

94-1. 



Heterogenous. 
Heterogenous. 
Heterogenous. 



270 POLLEN TETKAD SEGREGATION IN ASTROLOMA AND ACEOTRICHE. 

segregation similar to the model just described. Fragmentation does not occur in 
Grampians Ast. pmifoliuni or in Acrotriche fasciculiflora, and it must therefore con- 
stitute an additional complication rather than a primary cause of pollen tetrad 
segregation in the Styphelieae. 

Summary and Conclttsions. 

An analysis of pollen tetrad segregation in Astroloma pinifoHuvi and Acrotriche 
fasciculiflora has been attempted. 

In eastern Ast. pinifolium, segregations generally fit a trinomial square form, and 
occasionally fit the hypothesis of independence. Tetrad segregations in a population 
of the same species from south-west Victoria (Grampians) are very different. 

In Acrotriche -fasciculiflora, segregations occasionally fit the trinomial square form, 
but more often show significant departures from this form. 

The trinomial square form of tetrad frequency distribution requires strict 
equationality of the first meiotic division, and independence between the two second 
divisions in each mother cell, and a genetic or even chromosomal mechanism may be 
involved. 

Deviations from the trinomial square form could be determined by loss of strict 
equationality of the first division. In the case of genetic or chromosomal mechanisms 
this would require a shift in regions of chiasma localization. 

Consistent departures from expectations based on independence in Acrotriche 
fasciculiflora suggest that cytoplasmic conditions may be in part responsible for pollen 
death within tetrads. In particular, a gradient distribution of cytoplasmic substances 
may be Involved. 

Highly significant and very substantial intraplant variation in tetrad type 
frequencies occur, showing the behaviour to be sensitive to environmental modification. 
Until environmental effects can be evaluated, attempts to explain interplant and inter- 
population differences are hazardous. Acceptable model systems must permit substantial 
environmental modification. 

It is possible that both chromosomal and cytoplasmic systems operate as partial 
causes of segregation behaviour in the Styphelieae. 

Literature Cited. 
Kempthorne, O., 1957. — An Introduction to Genetic Statistics. John AViley & Sons. Inc., N. 

York. 
Levan, a., 1942. — A gene for the remaining in tetrads of ripe pollen in Petunia. Hereditas, 

28: 429-435. 
Martin, P. G., and Peacock,, "V^". J., 1959. — Pollen tetrad patterns in Leschenauttia. Proc. Linn. 

See. N.S.W., 84: 271-277. 
Smith-White, S., 1948. — Polarised segreg-ation in a stable triploid. Heredity, 2: 119-129. 
■ , 1955. — Chromosome Numbers and Pollen Types in the Epacridaceae. Aiist. J. Bot., 

3: 48-67. 
, 1959. — Pollen Development Patterns in the Epacridaceae. Pres. Address. Proc. 



Linn. See. N.S.W., 84: 8-35. 



271 



POLLEN TETRAD PATTERNS IN LESCHENAVLTIA. 

By P. G. Martin' and W. J. Peacock." 

(Plate X.) 

[Read 26tli AuKUSt, 1959.] 



Synopsis. 

Mature pollen grains of Leschenaultia remain associated in tetrads. In eleven collections 
from both wild and cultivated plants of three species, aborted pollen grains, varying in 
frequency from 7% to 53%, have been observed. All combinations of good and aborted grains 
in a tetrad were present in every collection. Univalent formation at meiosis has been 
observed. An hypothesis relating univalent distribution and pollen abortion patterns is 
presented. This, almost certainly, accounts for a large proportion of pollen abortion, but 
there may be other contributory causes. 



Introduction. 

The genus Leschenaultia R.Br, is different from other genera of the family 
Goodeniaceae in that the mature pollen grains remain in tetrads. This is uncommon, 
but not rare, among Angiosperms, but, whereas there is a tetrahedral arrangement of 
pollen grains in most such genera, in Leschenaultia there is a rhomboidal arrangement 
in a single plane. Mature pollen of Leschenaultia was stained with a cytoplasmic stain 
and it was observed that there was always a proportion of unstained, aborted pollen 
grains. Within tetrads every possible combination of stained and unstained pollen was 
observed, viz: 

Full tetrads . . 4 stained 

Triads . . . . 3 stained, 1 aborted 

Dyads A . . . . 2 stained, 2 aborted, the latter being adjacent 

Dyads O . . . . 2 stained, 2 aborted, the latter being opposite 
'^ Monads . . . . 1 stained, 3 aborted. 

Nullads . . . . 4 aborted 

These are illustrated in Plate x, figs. 1 and 2. 

This paper is concerned with observations of the frequencies of the different types 
and a consideration of hypotheses to explain the observations. 

Materials and Methods. 

Samples of mature pollen from plants of L. formosa R.Br, and L. linarioides DC. 
■Were collected by Mr. D. E. Symon in their native habitat in Western Australia. Most 
observations were carried out on plants of L. formosa and L. hiloha Lindl. grown in 
cultivation. The samples of L. linarioides and L. formosa collected in their native 
habitat were diploid (2n = 18). Among cultivated plants of both L. formosa and L. 
hiloha, diploids (2n = 18) and tetraploids were found. All the observations reported 
refer to diploid plants. It is worth recording that, whereas nearly all pollen produced 
by the tetraploid L. formosa plant aborted, the tetraploid L. hiloha plant produced 
approximately 70% "fertile" (i.e. normally stained) pollen. 

For mature pollen observations, buds were collected just before anther dehiscence 
and, after opening the corolla, fixed in acetic-alcohol (1:3). Anthers were dissected in 
a dextrin-sorbitol fluid similar to that described by Zirkle (1940) in which acid fuchsin 
replaced carmine. Aborted pollen grains were quite colourless and those assumed to be 
fertile stained bright red so that there was no doubt in distinguishing them. Little 

' Zoology Department, University of Adelaide, South Australia. 
'' Botany Department, University of Sydney, New South "Wales. 

Proceedings of the Linnean Society of New South Wales, 1959, Vol. Ixxxiv, Part 2. 



272 



POLLEN TETRAD PATTERNS IN LESCHENAULTIA, 



pollen was lost in mounting and tetrads were rarely broken or lying on their sides. 
All tetrads on a slide were counted and it is thought that errors in counting were 
negligible. 

For observations of meiosis, buds were fixed in acetic-alcohol (1:3) and sub- 
sequently stored in 70% alcohol. Following hydration, buds were hydrolysed for five 
minutes in N HCl at 60°C. and then squashed in 2% synthetic orcein (G. T. Gurr, 
London) in 60% acetic acid. The pre-treatment with HCl not only improves the stain but 
helps separate the pollen mother cells, which otherwise tend to adhere to the tapetum. 
Alternatively an aceto-carmine squash technique was used without previous hydrolysis. 

Results. 
1. Observations on Pollen Grains. 

Frequencies of the different types of tetrads and the frequencies of aborted pollen 
grains are shown in Table 1. 

Table 1. 
Frequencies of Tetrad Types in Various Collections. 



Species. 


Plant and Collection. 


03.- 


4 


1 


< 


d 

"i 
>> 


o 




Tetrads with 
at Least 
One Aborted 
Pollen Grain. 


Percentage 
Aborted 
Pollen 
Grains. 


L. linarioides . . 


Wild plant 1 . . 




1,652 


0-815 


0-073 


0-064 


0-012 


0-027 


0-009 


0-185 


12-1 


L. biloba 


Cultivated plant 1, 


Col- 






















lection 1 




297 


094 


0-189 


0-306 


0-037 


0-239 


0-135 


0-906 


53-3 




Cultivated plant 1, 


Col- 






















lection 2 




349 


0-185 


0-421 


0-248 


0-058 


0-072 


016 


0-815 


40-3 




Cultivated plant 1, 


Col- 






















lection 3 




2,345 


0-554 


0-258 


0-116 


0-013 


0-027 


0-032 


0-446 


18-1 


L. formosa 


WUd plant 1 . . 




1,405 


0-809 


0-122 


0-037 


0-009 


0-008 


0-015 


0-191 


7-4 




Cultivated plant 1 




970 


0-911 


0-041 


0-022 


0-014 


0-003 


0-009 


0-089 


3-9 




Cultivated plant 2 




1,658 


0-589 


0-293 


0-078 


0-014 


0-013 


0-013 


0-411 


14-2 




Cultivated plant 3 




703 


0-341 


0-404 


0-159 


0-033 


0-037 


0-026 


0-659 


24-7 




Cultivated plant 4 




934 


0-358 


0-442 


0-140 


023 


• 022 


0-015 


0-642 


27-4 




Cultivated plant 5, 


Col- 






















lection 1 




933 


0-250 


0-399 


0-260 


0-024 


0-044 


0-023 


0-750 


29-6 




Cultivated plant 5, 


Col- 






















lection 2 




811 


0-361 


0-4.30 


0-136 


0-023 


0-035 


0-015 


0-639 


22-8 



Although there is considerable variation in their frequencies, abortion patterns 
are similar in all three species whether collected in cultivation or the native state. Thus 
the underlying causes of abortion may be similar in all three species and must be 
present in the native state. 

The great variability between the three different collections from the one plant of 
L. diloba suggests that the environment has an important effect. This is confirmed 
by the two collections from the fifth cultivated plant of L. formosa. These were 
statistically different (x^ for heterogeneity = 51-818 with 3 degrees of freedom. P < -001). 

However, within a bud the contents of anthers are not heterogeneous. The 
individual counts for the five anthers from the one bud of L. formosa wild plant 1 are 
shown in Table 2. 

This result has been confirmed using buds from L. formosa cultivated plants 3, 4 
and 5, the probabilities derived from the contingency x^s being 10-20%, 10-20% and 
60-70% respectively. 

These data may be summarized as showing that there may be a large variation 
between buds on the same plant, but that the anthers within a bud are homogeneous. 

The large variation within a plant made it difficult to correlate observations on 
meiosis and on pollen grains. In the absence of facilities for growing plants under 
constant environmental conditions, collections of meiotic material of L. Mloba were 



BY P. G. MARTIN AND W. J. PEACOCK. 



273 



made and other buds, of similar length and presumably at the same stage, were 
marked for later pollen grain studies. Only one pair of collections was successful at 
both stages. The pollen grain counts are shown in Table 3. The three buds were not 
heterogeneous, which lends support to the assumption that they were at the same 
stage when marked. 

In some tetrads one or more small aborted microcytes were observed (Plate x, 
ng. 3). 



Numbers of Tetrad Types 



Table 2. 
in One Bud of L. formosa {Wild Plant 1). 



Anther. 


FuU 
Tetrads. 


Triads. 


Dyads A. 


Dyads 0. 


Monads. 


Nullads. 


Total. 


1 


239 


41 


10 


2 


2 


3 


297 


2 


219 


38 


7 


3 


1 


2 


270 


3 


219 


35 


16 





2 


8 


280 


4 


225 


27 


14 


7 


4 


5 


282 


5 


234 


31 


5 


1 


2 


3 


276 


Total 


1,136 


172 


52 


13 


11 


21 


1,405 



x'' heterogeneity=19-836 with 12 degrees of freedom* 010>P>005. 
* Dyads 0, Monads and NuUads were grouped into one class. 



Table 3. 
Numbers of Tetrad Types in Three Buds of L. biloba (Plant 1, Collection 3) Judged to have been at the Same Stage of 

Development. 



Bud. 


Eull 
Tetrads. 


Triads. 


Dyads A. 


Dyads O. 


Monads. 


NuUads. 


Total. 


1 


465 


247 


103 


14 


27 


26 


882 


2 


439 


185 


78 


8 


14 


21 


745 


3 


396 


172 


90 


9 


22 


29 


718 


Total 


1,300 


604 


271 


31 


63 


76 


2,345 



X* heterogeneity = 13 -233 with 10 degrees of freedom 0-20<P<0-30. 
2. Observations on Ponen Mother Cells. 

In all three species the diploid complement of chromosomes was 18. Detailed 
studies of meiosis were confined to L. forynosa and L. biloba. Prophase stages were 
difficult to observe. At metaphase I, a large proportion of bivalents was seen to have 
a single terminalized chiasma (Plate x, fig. 4). In L. biloba the chiasma frequency in 
one sample was 9-85 ± 0-95 and in a sample from L. formosa 11-55 ± 0-3. 

Univalents were frequently observed at metaphase I, anaphase I and early telophase 
I (Plate X, fig. 5). The frequencies for L. biloba (cultivated from plant 1) are shown 
in Table 4. 

Of 780 cells scored, 100 contained univalents, 98 cells with one pair and two cells 
with two pairs. Similar observations in three collections from L. formosa showed 
univalents to be present in 10-6%, 17-4% and 11-4% of pollen mother cells respectively. 

Collection 3 in Table 4 corresponds to the pollen grain data shown in Table 3, i.e. 
the buds in Table 3 were judged to have been undergoing meiosis at the same time as 
Collection 3, Table 5, was made. The important point to be noticed is that, whereas 
univalents appeared in only 23% of pollen mother cells, aborted pollen grains appeared 
in 44-6% of tetrads. 

Although frequencies have not been estimated, laggards have also been observed 
at telophase II (Plate x, fig. 6). It was also observed in cells at the second division of 
meiosis that spindles were always parallel and never crossed. An important implication 
of this is that, in a dyad-0 tetrad, a stained and an aborted cell must be sister cells 



274 



POLLE^^ TETRAD PATTERNS IN LESCHENAULTIA, 



from the same second division spindle. "When each spindle gives rise to one stained and 
one aborted pollen grain, usually equal numbers of dyad— A and dyad-0 tetrads should 
result. Dyad-A, but not dyad-0, tetrads could also result from some polar inequality 
at the first division of meiosis. 

Table 4. 
Number of Pollen Mother Cells of L. biloba Containing Univalents at Metaphase 1 — 

Telophase 1. 



Collection. 


Cells with 
Univalents. 


Total Cells. 


Percentage of 
Cells with 
Univalents. 


Collection 1 — 
Bud 1 
Bud 2 
Collection 2 
Collection 3 


27* 
18 
20* 
35 


273 
161 
195 
151 


10 
11 
10 
23 


Total . . 


100 


780 


— 


* Includes one cell \v 


th four univalents. 

Discussion 







The data in Tables 1, 2 and 3 show that, when buds, from one plant, at different 
stages of development were studied, the tetrad patterns were heterogeneous. However, 
when buds at approximately the same stage of development were studied, the tetrad 
patterns were not heterogeneous. Because anthers in a bud also were homogeneous it 
is deduced that the environment of the plant is important in determining the frequency 
of pollen abortion. 

One simple hypothesis which should be considered to account for the abortion 
patterns observed is that abortion was caused by a randomly occurring event. If this 
were so, the frequencies of the different patterns should have followed a binomial 
distribution. Thus in the data shown in Table 2, the frequency of aborted grains was 

414 

=-0744. The expansion of (-9256 + •0744)* should give the expectations in the 

5620 

different classes of tetrads. These are as follows. 





Full Tetrads. 


Triads. 


Dyads. 


Monads. 


Nullads. 


Expected 
Observed 


1030-4 
11.36 


332-2 
172 


40-2 

65 


21 
11 


Very small 
21 



Also the frequency of adjacent dyads should have been twice that of opposite 
dyads. The observations were 52:13. There were significant departures from both 
expectations and these were also shown by all other data. On its own, at least, this 
hypothesis must be discarded. 

The high frequency of univalents observed at meiosis suggests that pollen abortion 
might be explained by their distribution. The low frequency of cells with more than 
one pair of univalents (observed twice in 780 cells) suggests that an hypothesis could 
be formulated taking account only of cells with one pair of univalents. 

Let p be the frequency of formation of one pair of univalents; i.e. 1-p is the 
frequency of non-formation of univalents. Any one univalent may have one of three 
fates, i.e. : 

(1) Be included undivided at one pole at anaphase I. Let the frequency of this 
be X. 



BY P. G. MAKTIN AND W. J. PEACOCK. 



275 



(2) Be lost, i.e. not included at either pole at anaphase I. Let the frequency of 
this be y. 

(3) Divide at anaphase I, one chromatid going towards each pole. Let the 
frequency of this be z. 

Misdivision may be included with loss since the result would be the same. Thus 
xfy + z = l Equation 1. 

When a univalent divides, an individual chromatid may be lost, either before 
telophase I or during the second meiotic division. Let the frequency of this, for any 
one chromatid, be w. 

Using these parameters, the frequencies of the different tetrad types have been 
worked out and are shown in Tables 5 and 6. Table 5 is based on the additional 
hypothesis that deficiency only leads to abortion, i.e. disomic pollen grains are viable. 
Table 6 is based on the alternative hypothesis that any unbalance leads to abortion. 

The six equations derived from either table, together with Equation 1, could, in 
theory, be solved by the method of maximum likelihood (Mather, 1951, page 47) for 
any particular set of data. The task is, however, formidable and has not been achieved. 

Table 5. 

Tetrad Types Resulting from Distribution of Univalents on the Hypothesis that Only Deficiency Leads to Abortion. 

For further explanation, see text. To distinguish the two poles, they are called N and S. For brevity. " full tetrads " 

are shown as " tetrads ". 





\ 
\ 1st 

\ uni- 
\ valent 

2nd \ 
uni- \ 
valent \ . 


Lost. 


Included Whole. 


Divided and Included. 


Frequency. 




N. 


S. 


N" & S. 


N. 


S. 


Nil. 




Lost 


NuUad 


Dyad A 


Dyad A 


\ Dyad A 
i Dyad 


Monad 


Monad 


Nullad 


y 


Included 


N 


Dyad A 


Dyad A 


Tetrad 


Triad 


Dyad A 


Triad 


Dyad A 


X 


whole 


S 


Dyad A 


Tetrad 


Dyad A 


Triad 


Triad 


Dyad A 


Dyad A 


X 

2^ 




N & S 


i Dyad A 
i Dyad 


Triad 


Triad 


i Tetrad 
i Triad 
J Dyad A 
\ Dyad 


i Triad 
i Dyad A 
i Dyad 


4 Triad 
i Dyad A 
i Dyad 


J Dyad A 
1 Dyad 


zw^ 


Divided 

and 
included 


N 


Monad 


Dyad A 


Triad 


i Triad 
i Dyad A 
i Dyad 


\ Dyad A 
* Dyad 


J Dyad A 
J Dyad O 


Monad 


zw(l— w) 




S 


Monad 


Triad 


Dyad A 


i Triad 
i Dyad A 
i Dyad 


i Dyad A 
* Dyad 


J Monad 
h Dyad A 


Monad 


zw(l— w) 




Nil 


Nullad 


Dyad A 


Dyad A 


k Dyad A 
J Dyad 


Monad 


Monad 


Nullad 


z(l-w)' 




Frequency 


y 


X 

2" 


X 


zw' 


zw(l— w) 


zw(l— w) 


z(l-w)= 





Total frequencies : 

Full tetrads = (l-p)+p(ix^ + iz=w'). 
Triads =pzw[2x + zw-(2 — IJw)]. 

Dyada A =px[2y +2z(l-w) + Jx] +pzw={y + ^[iw«^-(l-w)(3- 
Dyads =pzw2[y+izw^+z(l— w)(2— w)]. 
Monads =pzw(l— w)[4y 4-z(l— w)(4 — 3w)]. 
NuUads =py=+pz(l-w)''[2y + z(l-wr-]. 



■2w)]}. 



276 



POLLEN TETKAD PATTERNS IN LESCHENAULTIA, 



If it were done the whole univalent hypothesis and the alternative subsidiary hypotheses 
could be tested. 

Observations of univalents at meiosis I, of laggards at meiosis II and of microcytcs 
within spores showing abortion patterns support the univalent hypothesis. Microcytes 
presumably form around the univalents. The frequency of tetrads containing one or 
more aborted spores should be less than the frequency of pollen mother cells containing 
univalents because a full tetrad may be formed even when univalents are present at 
meiosis (Tables 5 and 6). In the observation reported here, the frequency of tetraas 
containing one or more aborted spores ranged from 9% to 90% with a mean of 52%; 

Table 6. 

Tetrad Types Resulting from. Distribution of Univalents on the Hypothesis that any Unbalance Leads to Abortion. 

For further explanation, see text. To distinguish the two poles, they are called N and, S. For brevity, " full tetrads " 

are shown as " tetrads ". 





\ 1st 
\ uni- 
\ valent 

2nd \ 
uni- \ 
valent \ 


Lost. 


Included Whole. 


Divided and Included. 


Frequency. 




N. 


S. 


N & S. 


N. 


S. 


Nil. 




Lost 


Nullad 


Dyad A 


Dyad A 


i Dyad A 
i Dyad 


Monad 


Monad 


NuUad 


y 


Included 


N 


Dyad A 


Nullad 


Tetrad 


i Dyad A 
i Dyad 


Monad 


Triad 


Dyad A 


X 

2 


whole 


S 


Dyad A 


Tetrad 


Nullad 


i Dyad A 
i Dyad 


Triad 


Monad 


Dyad A 


X 

2 




N&S 


i Dyad A 
i Dyad O 


i Dyad A 
i Dyad 


* Dyad A 
J Dyad 


J Tetrad 
i Nullad 
i Dyad A 


* Monad 
i Triad 


i Triad 
i Monad 


i Dyad A 
i Dyad 


zw* 


Divided 

and 
included 


N 


Monad 


Monad 


Triad 


i Monad 
I Triad 


i Dyad A 
i Nullad 


i Dyad A 
i Dyad 


Monad 


zw(l— w) 


S 


Monad 


Triad 


Monad 


i Triad 
i Monad 


* Dyad A 
i Dyad 


J Dyad A 
i Nullad 


Monad 


zw(l— w) 




Nil 


Nullad 


Dyad A 


Dyad A 


i Dyad A 
* Dyad 


Monad 


Monad 


Nullad 


Z(l-W)2 




Frequency 


y 


X 

~2 


X 


zw^ 


zw(l — w) 


zw(l— w) 


Z(l-W)2 





Total frequencies : 

Full tetrads = (l-p)+p(ix^ + iz=w'). 

Triads =2pzw(l— wXx + zw''). 

Dyads A =pzw2[x+y+z(iw2+3{l-w)2]+2px[y + z(l-w)2]. 

Dyads O =pzw^[x+y+2z(l-w)=']. 

Monads =2pzw(l-w){x+2y+z[w2+2(l-w)2]}. 

Nullads =pz(l-w)2{2y + z[w=-(-(l-w)^]}+p(y2+izw2+Jx^). 

the frequency of pollen mother cells containing univalents ranged from 10% to 23% 
with a mean of 13%. Although there is a considerable difference between these mean 
values, this cannot be taken as disproof of the univalent hypothesis because of the great 
variability and the overlapping distributions. Moreover, with one exception, there 
was no direct correlation between the two sorts of observations. In the one set of data 
in which an attempt was made to estimate frequencies of tetrad types which could bo 
correlated with meiotic observations, tetrads with aborted spores were twice as frequent 
as pollen mother cells with univalents. This suggests that univalent distribution alone 



BY P. G. MARTIN AND W. J. PEACOCK. 277 

can account for only part of the observed abortion, but it is not a rigorous enough test 
to exclude its accounting for all abortion. 

If univalent distribution cannot account for all abortion, there must be other 
causes co-existing, e.g., abortion due to randomly occurring events would increase the 
frequency of triads and dyads, but would make less difference to the frequencies of 
monads and nullads. A further hypothesis which should be mentioned is that proposed 
for some species (e.g., Astroloma pinifoUum) of the Epacridaceae by Smith-White in 
a paper in this Part of these Proceedings (pp. 259-270). If, at meiosis, the segregation 
of cytoplasmic and/or chromosomal factors is equational at the first division but reduc- 
tional at the second, such as to cause abortion at both poles, one pole or neither in each 
half mother cell in the ratio x:y:z, then the different tetrad types can be derived 
from the trinomial (x + y + z)l Because univalent distribution is responsible, almost 
certainly, for a large proportion of abortion, no attempt has been made to apply the 
trinomial hypothesis alone. It is mentioned as a possible co-existing cause of abortion, 
although it is considered unlikely to be operating because terminal localization of 
chiasmata causes reduction at the first meiotic division for most chromosomal material. 

It is not apparent how a rather high proportion of inviable gametes persists in 
these species. Presumably this selective disadvantage is outweighed by some contingent 
advantage. For example, the terminal localization of a single chiasma may preserve 
intact different combinations of genes, the presence of which is advantageous to a local 
population of species. The chance failure of formation of this chiasma would result 
in univalents and relative infertility. Possibly the disadvantage of this is not as great 
as the advantage of terminal localization of a single chiasma. This may be particularly 
true where there is always some fertile pollen and pollination is very efficient as it 
may be in Leschenaultia with its elaborate indusium for collecting and holding pollen. 

Acknowledgements. 
We would like to thank Dr. S. Smith-White for valuable advice and criticism. One 
of us (W.J. P.) carried out this work while holding a C.S.I.R.O. post-graduate 
studentship. 

References. 
Mather, K., 1951. — The measurement of linkage in Heredity. Methuen & Co. Ltd., London. 
Smith-White^ S., 1959. — Pollen Tetrad Segregation in Astroloma pinifoUum and In Acrotriche 

fasciculifiora. Proc. Linn. See. N.S.W., 84 : 259-270. 
ZiRKLB, C, 1940. — Combined fixing, staining and mounting media. Stain Tech., 13: 139-153. 

EXPLANATION OP PLATE X. 

Fig. 1. Camera lucida drawings of the various types of tetrads in Leschenaultia formosa. 
Top (left to right) : full tetrad, triad, dyad O ; bottom (left to right) : nullad, dyad O, monad, 
dyad A. x 200. 

Fig. 2. ' Leschenaultia formosa. Pollen tetrad's stained with acid fuchsin showing full 
tetrads, triad, dyad .\. and monad, x 100. 

Fig. 3. Leschenaultia hiloba. Pollen tetrads showing full tetrads, dyad O and dyad A, 
with one microcyte. x 100. 

Fig. 4. Leschenaultia biloba. Pollen mother cells at metaphase I. In the majority of 
bivalents there is a single, terminally localized chiasma. Aceto-orcein ; x 830. 

Fig. 5. Leschenaultia hiloba. Pollen mother cell at telophase I, showing two univalents. 
Aceto-orcein ; x 830. 

Fig. 6. Leschenaultia formosa. Pollen mother cell at telophase II showing laggards. 
Aceto-carmine ; x 830. 



278 



SEED COAT ANATOMY AND TAXONOMY IN EUCALYPTUS. II.* 

By E. Gauba and L. D. Pryor. 

(With Plates xi-xiii; Text-figures 20-39.) 

[Read 26th August, 19.59.] 



Synopsis. 

Blakely's Series Eudesmiae of the Section Macrantherae comprises 12 species of which 
1 1 have been examined. Except the Subseries Leptospermae their seed's are of hemitropouE 
structure, bifacial, the dorsal side uniform, the ventral side containing chalaza (with the 
vascular system), hilum and micropyle. There is an immediate transition of the conducting 
tissue from tlie placenta to the chalaza, that is to say without the formation of a raphe. 
The inner epidermis of the outer integument is a crystal epithelium par excellence, the inner 
integument is suberized, the inner cuticle very tender and smooth (or with inconspicuous 
projections). 

Apart from these common features the separate species often display considerable 
differences not only in morphological aspects (seed smooth, ribbed' or winged) but particularly 
with regard to the anatomy of epidermis, hilum, micropyle etc., suggesting some necessary 
rearrangement within the Eudesmiae or even a transfer of some species to other groups of 
the Macrantherae (or the reverse). 

The Subseries Leptospermae (E. tenuipes and E. cu7-tisU) exhibits — as far as the testa 
structure is concerned — no relation to the hemitropous Eudesmiae. The seeds are anatropous 
with a typical raphe, hence with a different relative position of chalaza, hilum and micropyle 
and a different vascularization. The crystal epithelium is missing, the inner cuticle rather 
conspicuous and with relatively prominent ribs. There is no doubt from the complete uniformity 
in testa structure that these two species are very close relatives and sharply demarcated from 
the remainder of the Eudesmiae. 



Section MACRANTHERAE. Series Eudesmiae (Benth.) Blakely. 

Introduction. 
Of the twelve species listed by Blakely in the Series Eudesmiae the seed of the 
following eleven were examined. 

(1) Subseries Fasciculares: E. erythrocorys. E. tetragona, E. eudesmoides, E. 

ebbanoensis. 
Subseries Ef asciculares : E. odontocarpa, E. tetrodonta. 
Subseries Holocalyces: E. lirata, E. similis, E. baileyana. 
Their seeds are of hemitropous structure and we distinguish between the ventral 
side (with chalaza and its vascular system, hilum and micropyle) and the uniform 
dorsal side. 

(2) Subseries Leptospermae: E. tenuipes and E. curtisii. 

Their seeds are of anatropous origin and we may distinguish between a raphe side 
(with chalaza and raphe bundle) and a uniform raphe-free side. Hilum and micropyle 
lie side by side at the basal end of the seed. 

The following explains the lettering in the Text-flgures and Plates: cd, clinging 
disc; ch, chalaza; cot, cotyledons; c?; crystal epithelium; ec, embryo cuticle; end. 
endosperm; h, hilum; hy, hypocotyl; ic, inner cuticle; ie, inner epidermis; ii, inner 
integument; Ic, lignified cells; m, micropyle; mu, mucilage; nr, nucellus remnants; oc, 
outer cuticle; oe, outer epidermis; oi, outer integument; rm, root meristem; vb, vascular 
bundle. 

In the following the different parts of the testa are described together with some 
remarks on the nucellus, endosperm and embryo. 

Outer Integument. 
Originally a transitory storage tissue supplying the growing seed with toed 
material, it undergoes during maturation various alterations in contents and structure 
of its cells. The final stage which is discussed here is that reached in ripe abscissed 
seed. 

* See Part I, Proc. Linn. Soc. N.S.W., 1958, 83, Part 1 : 20-32. 
Proceedings of the Linnean Society of New South Wales, 1959, Vol. Ixxxiv, Part 2. 



BY E. GAUBA AND L. U. PRYOK. 279 

Outer EpMermis. 

The somewhat tabular cells are polygonal In surface view and more or less iso- 
diametric (Fig. 20, 21), the only exception being E. erythrocoryfi where most of the 
cells are elongated parallel with the surface < Fig. 22). The outer walls are thin, giving 
cellulose reaction (E. eudesmoides, E. 'baileyana. E nimilis) , or thick and lignified 
(E. erythrocorys) or have mucilage as secondary wall deposits (E. tetragona. E. 
ehbanoensis) . The mucilaginous cells of E. odontocarpa are of particular interest 
because of the complex structure of the outer wall. The smaller cells have a 
mucilaginous layer beneath a thin outer wall of cellulose, outlined against the lumen 
by a tender cellulose membrane (Fig. 23a). In the large bulging cells there is next 
to the lumen a thick, lignified and pitted membrane (Fig. 23?;). In some instances 
this membrane was seen embedded within the mucilage of which the inner layer 
showed traces of pits continuous with those of the lignified membrane (Fig. 23c). This 
suggests that a formerly solid wall underwent it^'o parte a subsequent mucilaginous 
modification. In E. tenuipes and E. curtisii not only all epidermal cells have membrane 
mucilage but also many cells of the integumentary parenchyma (Fig. 24). The mucilage 
is impregnated with brown tannin-like substances. After their removal with Eau de 
.Javelle it stains blue with chloroiodide of zinc. Cuprammonia stains for a short while 
light blue, whereupon, swelling rapidly, the mucilage exhibits striking lamellations 
before being dissolved. 

Phlobaphene deposits in epidermis cells are a regular feature in the anatropous 
Budesmiae (E. tenuipes, E. curtisii), recalling the anatropous Renantherae. 

Where the outer walls are thin they have collapsed in ripe seed and are more or 
less appressed to the inner walls rendering the. surface minutely pitted (E. haileyana, 
E. similis). On the other hand a gregarious occurrence of thick-walled bulging cells 
causes the wrinkled appearance (E. odontocarpa). In E. tetrodonta the sometimes 
warty surface is due to groups of lignified cells elongated perpendicularly to it. 

The cuticle covering this epidermis is well preserved in E. erythrocorys and E. 
tenuipes, but in E. baileyana the usual stains and reagents failed to reveal its presence. 
In some sections of E. eudesmoides (and some other species) the cuticle was seen as 
an unbroken line, in others only in fragments, more or less detached from the 
epidermis, or, if not entirely missing, merely as a dotted line. Of course, the ovule 
always has a perfect cuticle which during maturation to seed may become disorganized 
or even completely resorbed. A loosening of its contact with the epidermis may also 
occur so that sectioning followed by treatment with reagents may tear off and wash 
away what remained of it. 

Inner Epidermis. 

This epidermis is, in all seeds of hemitropous structure, a typical crystal epithelium, 
each cell containing one large and a few smaller crystals of monoclinic calcium 
oxalate monohydrate. Crystalline granules (crystal-sand) are also often present, some- 
times also elongated crystals (styloids). 

The most common well-developed crystals are six or eight-faced, more rarely ten- 
faced. Professor Machatschki* (in lift.) gives the following indices which we reproduce 
with his sketches (Text-fig. 25). 

Hence we may infer that Figure 25o is a combination of the prism (110) and the 
basal pinacoid (001), Figure 25& shows an additional clinopinacoid (010) and finally 
the ten-faced crystal in Figure 25c also the orthopinacoid (100). 

The larger iTystals are encased within cellulose sheaths which, after dissolving 
the crystals, imitate so precisely their outline that only with a polarization microscope 
are these structures recognized as merely empty forms. 



* We are greatly indebted to Prof. Dr. F. Machatschki, Director of the Institute of 
Mineralogy at the University of Vienna, for the examination of crystals in the seed coat of 
Fi. eiidesvioides. 



280 



SEED COAT ANATOMY AND TAXOKOMY IN EUCALYPTUS. II, 



The cells are closely packed, their morphologic outer walls thick, constricting the 
lumen to such extent that there remains very little room to accommodate the crystals 
(Pig. 28, 32). This may explain their tabular habit parallel to the basal pinacoid 
(001). 

As part of the integument which develops from a more or less circular rim at the 
base of the future nucellus the crystal layer of the mature seed exhibits at its base a 
circular, elliptic or oval gap, the "window in the crystal armour" — a very appropriate 
designation. 




Ve ic 



lOO 




no 




oio 



no 



16 



Text-fig. 20-24. — 20-22, surface visw of the dorsal epidermis, lOOx. 20, E. baileyana: 
21, E. tetrodonta, showing a group of sclerified cells; 22, E. erythrocorys. 23-24, mucilaginous 
epidermis cells. 23a-c, E. odontocarpa, three types of outer wall modifications, 160x ; 24, E. 
tenuipes, transv. sect, through an edge of seed. Mucilage dotted, phlobaphenes black, 2 5 Ox 

Text-fig. 25a-c. — E. eudesmoides, oxalate crystals of the inner epidei-mis (Crystal 
epithelium) of the outer integument. 



The cells of this epidermis are crystal-bearing to its very fringe {E. tetragona. E. 
eudesmoides, E. 'baileyana) or some rows bordering the "window" are empty (E. 
odontocarpa) and moreover both empty and suberized in E. erythrocorys and E. 
ebhanoensis . The border is more or less smooth or as in E. similis jagged (Fig. 
35?). c). 

This striking crystal epithelium in hemitropous seed is missing in the anatropous 
Subseries of Leptospermae (E. tenuipes and E. curtisii) . Crystals occur here, too, but 
only dispersed in the hilar and chalazal region and along the raphe, the cells having 



BY E. GAUBA AND L. D. PKYOK. 



281 



only slightly thickened outer walls (Fig. 24 ie). 
epidermis has no crystals. 



The far greater part of the inner 



Integumentary Parenchyma. 

The parenchymatous tissue between the two epidermal layers and originating from 
them is always multi-layered on the ventral or raphe side. From here the number of 
layers gradually decreases towards the opposite side where it is reduced to a few {E. 
erythrocorys, E. tetragona) or to a single, often interrupted one (as in most of the 
species), and sometimes even missing {E. baileyana, E. lirata). A greater amount of 
this parenchyma is also found in wings (Fig. 27, 28) and in the costae. 




■ Text-fig. 2G-28. — Transv. sect, through ridges and wings. 26, 
27, E. tetragona, wing-, 130x ; 28, ,E. odontooarpa, wing, 250x. 



B. eyytlirocorijs, ridge, 50x ; 



The homogeneity of this tissue is often disturbed by the occurrence of cells — single 
or in groups — with different contents or wall structure. Crystal idioblasts are for instance 
very frequent in the ventral side of E. ehbanoensis, with crystals partly embedded in 
the uneven wall thickenings (PI. xi, fig. 5b), in the mlcropylar region of E. erythrocorys 
(PI. xi, fig. la), in wings of E. tetragona, etc. Sclereid idioblasts are not infrequent in 
E- erythrocorys, in E. tetrodonta adjacent to the inner epidermis and in E. ehbanoensis 
among suberized cells below the hilum (PI. xi, fig. 5b). In E. erythrocorys suberized 
cells along the bundle branches were observed. Mucilaginous cells occur in E. tenuipes 
(Fig. 24) and E. curtisii. Phlobaphene deposits are often seen (E. tetrodonta, E. 
tetragona) and impregnation of all walls with tannins, including those of both epidermal 
layers, is the rule. Because of this the seed coat is of dark (brown to black) appearance. 
I 



282 seed coat anatomy and taxonoiiy in eucalyptus. ii, 

Inner Integument. 
This is always two-layered and suberized. The cells are tabular, without inter- 
cellular spaces, empty and very often completely obliterated. Like the outer integument 
it embraces with its circular base the chalaza. This is true also of its inner cuticle 
v/hich is always present though not as conspicuous as in the Renantherae.* However, 
the median cuticle, separating the two integuments and present in the ovule, is missing 
in mature seeds. Only in E. tetrodonta were some fragments of it seen. 

Ridges and Wings. 

The seeds of E. l)aileyana, E. siviilis, E. lirata and E. ebbanoensis have a smooth 
surface, that is to say their testa is not expanded into ridges or wings. 

The angular seeds of E. erythrocorys have prominent ridges built up by epidermis 
cells exceedingly elongated at right angle to the surface (Fig. 26). The ridges in 
E. tetrodonta are of similar structure. 

E. tetragona has wings, the broadest among the Eudesmiae, though scarcely 
exceeding 1| mm. in width. The cross section through the circumferential wing reveals 
that the outer epidermis and parenchyma of the outer integument are the constituent 
parts (Fig. 27). 

E. eudesmoides and E. odontocarpa have a circumferential membranous wing 
separating the ventral side from the dorsal. This delicate and narrow expansion has 
nevertheless a complex structure as both integuments and endosperm tissue (or in the 
vicinity of the chalaza, nucellar tissue) participate in its build up. The presence of 
the thick-walled crystal layer confers some mechanical protection which in E. odonto- 
carpa is increased by lignification of some epidermal cells (Fig. 28). During germination 
the wing at the root pole is neatly split by the pressure of the swelling clinging disc, 
T/hereupon the hypocotyl emerges between the wing halves. Therefore the role of the 
clinging disc is not only to fix the germinating seed to the substratum, but as it 
expands by growth to burst the testa first of all. 

HiLUM. 

After the shedding of the seed an elliptical, more or less discernible scar, the hilum, 
marks the abscission region. In hemitropous species it is located on the ventral side, 
overlapping partly the chalaza, though sometimes but to a very small extent (E. 
evAlesmoides). However, the vascular bundle always has direct access to the chalaza. 
Therefore, there is no "raphe" so characteristic of the anatropous seed (E. tenuipes 
and E. curtisii). 

In E. haileyana, E. tetragona. E. eudesmoides and E. similis the hilum is bordered 
by thin-walled peripherally elongated cells forming a sort of parenchymatous sheath 
(Fig. 29, 30). In E. erythrocorys the hilum is sharply circumscribed by a rim of 
sclereids (Fig. 31). 

The surface of the hilum is flat or a shallow depression {E. bailey ana, E. 
eudesmoides, etc.) or convex and shield-like (E. ebbanoensis, PI. xi, fig. 5b). 

Three types of cicatrization of the hilum were observed: 

(1) Desiccation, shrinkage and collapse of the exposed cells, giving the hilum a 
pulverulent appearance, as for instance in E. tetragona, E. eudesmoides, E. similis. 
E. erythrocorys (Fig. 31). There is no suberization of the exposed cells and even 
cutinization was not traceable. Therefore, this is a very primitive type of wound 
healing. 

(2) In E. odontocarpa the hilar scar is sealed by sclereids with lignified walls 
impregnated with tannins. These sclereids are either closely packed (PI. xii, fig. 1) or 
dispersed in groups, and belong not only to the uppermost layer but also to the 
subjacent tissue (Fig. 32). It must be pointed out, however, that the scleriflcation 
seldom extends over the entire hilum, but is mostly restricted to that part which lies 

* In some instances where this cuticle was very delicate and the usvial stains were not 
adequate we used successfully a Rosanilln solution discoloured with SOf>. 



BY E. GAUBA AND L. D. PRYOR. 



28J 





Text-fig. 2il-33 (liilar structures). — 29-30, surface view of the hilum (ft) overlapping the 
chalaza (ch). Some tissue partly removed to show the extent of the chalaza (broken line) 
and the vascularization. 29, E. baileyana, 50x ; 30, E. tetragona, 65x, three bundle strands 
enter separately into the chalaza. The arrow points in the direction of the micropyle m. 
31-32, trans, sect, through the hilum. 31, E. erythrocorys, lOOx. A rim of macrosclereid's 
surrounds the hilum in which a vascular bundle (vh) enters; 32, E. oAontocarpa, 160x. The 
hilum is sealed by sclereids (metaplasy by sclerification). 33, E. erythrocorys, surface view" 
of the hilum with dispersed sclereids, 200x. 



'284 SEED COAT ANATOMY AND TAXONOMY IN EUCALYPTUS. II, 

outside the chalaza (PI. xii, fig. 1). Thus the portion overlapping the chalaza was 
often seen without scleriflcation but sometimes bordered by a more or less complete 
rim of thick-walled cells. This form of cicatrization represents, at least partially, a 
case of metaplasy* by scleriflcation and wall impregnation by antiseptic substances. 

Sclereids with lignifled uneven wall thickenings can be seen also in the hilum of 
E. erytlirocorys where they occur singly or in groups, variably dispersed among thin- 
walled cells (Fig. 33). Their occurrence is too scanty to be of any value for the 
protection of the scar, but unequal tensions in such a heterogeneous tissue may 
facilitate the final separation of the seed. 

(3) The most perfect case of metaplasy, that is by suberization, is exhibited in 
E. ehbanoensis where all surface cells of the hilum are suberized and completely filled 
with solid phlobaphenes extremely resistant to acids and bases. Suberized cells occur 
also in the underlying layers of the hilar cushion, intermixed with crystal- and 
sclereid-idioblasts (PI. xi, fig. .5a-?;). This is a remarkable case because Netolitzky, 
who critically revised the entire literature on Angiosperm seeds (about 1,100 papers), 
states that hitherto (1926) no suberization has been observed on the hilum. Hintringer 
(1927), who examined 80 species of 9 families, confirms Netolitzky's statement. 

It is worth mentioning that the agents stimulating the formation of sclereids 
(E. odontocarpa) or inducing suberization {E. ehhanoensis) bring about the same 
metaplasic changes in some epidermis cells adjacent to the hilum (PI. xi, fig. 5a). 
Analogy can be drawn with the traumatic stimulus of wound hormones in tissues not 
directly affected by injuries. 

In the hilum of E. similis suberization is confined to a few cells surrounding the 
entrance of the bundles into the chalaza. 

As we are concerned only with the hilar structure of the mature abscissed seed 
the mechanism of its separation from the placenta and the time of the metaplasic wall 
modifications are not discussed. This would require ontogenetic investigations. But 
it seems quite certain that the metaplasy precedes the abscission. 

Finally, the break in the vascular bundle occurs in the transition region between 
placenta and hilum where short tracheids have been preformed (Fig. 31). The some- 
what protruding bundle stubs can be located on the hilum of many species. 

Chalaza. 

The crystal layer, the suberized inner integument and the inner cuticle form 
together at their base a "window" which embraces and delimits laterally in a spectacular 
way the inner part of the chalaza. The outer part (above the window) passes 
imperceptibly into the parenchyma of the outer integument. 

Tm^o features characterize the chalaza: (1) The presence of a suberized tissue, the 
so-called "chalaza cork", in mature seed closing the gap and filled with dark coloured 
phlobaphenes. Thus, in some instances the chalaza is externally discernible without 
optical aid as a dark spot, in particular in those cases where these deposits are confined 
Piot only to the suberized part but occur throughout the whole chalaza. (2) The 
presence of the conducting tissue above the suberized zone. 

As in the Renantherae the nucellus tissue! can participate in a variable amount in 
the formation of the chalaza cork. In E. baileyana and E. tetragona it originates 
almost entirely from the nucellus, in E. odontocarpa it is chiefly the true chalaza 
parenchyma of which the inner part undergoes the suberization. E. tetrodonta is 
exceptional in having no chalaza cork. 

* The term "metaplasy" was introduced into plant anatomy by Kiister (1903) to designate 
progressive changes in histological character other than by cell growth or division. In our 
case these changes concern chiefly a scleriflcation or a suberization to form a protective 
cicatrice over the abscission region. 

t In sections through the chalaza an imaginary line closing- the gap in the inner cuticle 
delines neatly the chalaza from the nucellus. 



by e. gauba and l. d. phyor. 285 

Vascularization. 

The vascular supply to floral parts has received a great deal of attention in recent 
years. It has been suggested that the vascular bundles of an organ are more con- 
servative than its external form and therefore they may contribute to the solution of 
phylogenetic and taxonomic problems. A critical review of the whole literature — pro 
and contra — -was given by V. Puri (1951). 

In the Eudesmiae with anatropous seed the vascular bundle is brought up from 
the hilum to the chalaza through the raphe (Fig. 34a, h). 

In the Eudesmiae with hemitropous seed the conducting tissue enters directly 
from the placenta into the chalaza. In some species the number of strands penetrating 
the seed is not constant. In E. tetragona we saw one to three (PI. xiii, fig. 2 and Text- 
fig. 30), in E. similis (Fig. 35a, h, c) up to four strands entering separately into the 
chalaza. A variable number can be seen also in E. erytlirocorys, whereas; in E. 
eudesmoides (PI. xii, fig. 2), E. ehbanoensis, E. haileyana (PI. xiii, fig. 1) and E. lirata 
(Fig. 36a, h) we observed always one single strand. 

The bundle entrance never lies in the centre of the hilum but always close to its 
upper border. From here the bundle spreads immediately into several branches {E. 
odontocarpa, PI. xii, fig. 1) or the ramification starts at a short distance from its 
entrance (E. baileyana, Fig. 29 and PI. xiii, fig. 1; E. tetrodonta, Fig. 37a, b). 

The whole venation system seldom lies within the chalaza, most frequently some 
of the stronger branches penetrate somewhat the parenchyma of the outer integument 
but without taking an intra-integumentary course. 

In E. odontocarpa the nearly straight branches spread palmately in the chalaza 
(PI. xii, flg. 1), whereas in other species they are cui-ved in varying degrees, displaying, 
especially the outer ones, an arctuous course along the chalaza border, as for instance 
in E. baileyana (PI. xiii, flg. 1). Even abrupt angular turns can be observed (Fig. 35). 

The bundles are amphicribral, the tracheary elements being only tracheids with 
helically thickened walls. Veins and veinlets terminate either freely with a single 
or double flle of tracheids, or some of the branches may merge at their tips, as for 
instance in E. baileyana (Fig. 29) or E. similis (Fig. 35). The latter has sometimes 
transverse anastomoses interconnecting main strands (Fig. 35c). 

Considered as a whole the vascular supply is very poor in E. tetragona (PI. xiii, 
flg. 2), whereas in E. erythrocorys it displays a high degree of ramification and, by 
merging of the branchlet endings, a reticulate venation. 

The venation patterns as shown here for hemitropous seeds exhibit a relatively 
wide range of variation within this small Series, even allowing for the variations 
within the same species. Thus at first glance it seems doubtful whether any taxonomic 
conclusion can be reached at this stage before the investigation is extended to the 
remaining hemitropous Macrantherae. 

MiCROPYLE. 

Its approximate position in hemitropous seeds is given by prolongation of the hilar 
axis towards the root pole. On this line it can be discerned externally, sometimes 
even with the unaided eye, as a small tubercle or callosity (E. odontocarpa. Fig. 38a; 
E. baileyana), or it is hidden between ridges ( B. erythrocorys, E. tetradonta) or wings 
(E. tetragona). 

Inner and outer integument participate in its build-up, forming respectively the 
endostome and exostome. Due to the uniformity of the inner integument the endostomc 
is of a relatively simple structure, whereas the exostome is bounded by three anatomic- 
ally different parts (outer epidermis, integumentary parenchyma, crystal layer), of 
which each can contribute in various degrees to its formation. Thus, the micropyle 
as a whole exhibits in some species a very complex structure. 

Reports upon the micropyle structure of mature seeds in general are rather scarce, 
owing probably to technical difficulties in obtaining suitable preparations. We have 
found the best way is to isolate in the micropylar region the inner cuticle, inner 



286 



S1CKI> COAT AXA'IOMY AVI) TAXOi\OJM\ IN KIJCALYPTUS. II, 




56 a -^ 





}6 b 




W b 



}5[> 



Text-fig. 34-37. — \'ascularization patterns. 3-1, anatropous species (Leptospermae). 3ia, 
JJ!. tenuipes, 85x ; 34&, E. cuTtisii, 96x. 35-37, hemitropous species. 35f(-c, E. siniiUs, 45x (a 
with one, b with two, c witli four separately entering bundle strands) ; Z6a-hj E. lirata, 96x ; 
?,l((-lt, E. tetrodonta, 45x. 

(Broken line: chalaza extent; dotted line: hilura.) 



BY E. GAUBA AND L. D. PKTOK. 



287 



integument, crystal layer and outer epidermis and to examine separately these layers 
in surface and side view. 

Without claiming completeness some of our observations are illustrated in Text- 
figure 38 and on Plate xi. Variations, sometimes considerable, are frequent within the 





Text-fig'. 38a-d. — Variation.s ol" the micropylar structure in I'J. odotitocari>a, ca. 160x. 
«, surface view; b, c, exostome aperture in the (isolated) outer opidermis ; d, endostoine 
aperture (d was lying- beneath c). 

Text-figr. 39a-e. — Seed diagrams showing on the ventral sid'e the relative position of 
chalaza (ch) with the vascular bundle (vh), the hilum (/i), the micropyle (■»!). the hypocotyl 
(hy) with root meristem (r?»!) and clinging disc (cd), and the cotyledons (cot), ca. 9x. 

a, E. baileyana ; b^ E. odontocarpa ; c, E. tetrodonta ; d, E. ebbanoensis ; e, E. eudesnioides. 

same species, especially with regard to the aperture of the micropyle. This is under- 
standable if we bear in mind that the integuments have to keep pace with the increase 
in volume of the growing embryo, but there are differences in the rate of their growth 
which determine the final shape and size of the aperture in the integuments respectively 
in their different layers. On the seed surface for instance the micropyle may appear 



288 SEED COAT ANATOMY AND TAXONOJIY IN EITCALYPTUS. II, 

as a circular or irregular opening (PL xi, fig. 2a), as a narrow or gaping slit (Pig. SSb, 
c), or it may be closed by contact or overlapping of its margins. 

In separating the different layers in the micropylar region we first remove the 
endosperm which vaults gently into the suberized endostome (PI. xi, fig. la) and is 
rich in oil and proteins. The inner cuticle is said in general to be apparently missing 
beneath the micropyle. But in E. tenuipes we could isolate it as a tender membrane with 
rib-like projections simulating a cellular structure (PI. xi, fig. 3&). 

The inner integument penetrates as a small or elongated hump (PL xi, fig. Ic, 2b), 
open or closed on the top, or as a two-lipped projection (the lips parted or firmly 
appressed) more or legs deep into the exostome. In hemitropous seeds the inner 
limit of the exostome is well defined by the crystal layer which can be more or less 
vaulted, closed or with an open slit, the bounding cells of which are frequently devoid 
of crystals (PL xi, fig. 41)). We remember tliat in some species a crystal-free border 
of the crystal epithelium surrounds the chalaza. 

In the exostome of E. tetragona the crystal layer is elongated into a tube or a 
horn-like protuberance which in E. erythrocorys can be up to 1 mm. long terminating 
just beneath the outer epidermis through which a capillary canal leads outwards (PL xi, 
fig. la). But in some instances we saw this "crystal horn" either as a blunt cone 
(PL xi, fig. 16) or attenuated into a fine point and covered with integumentary 
parenchyma protruding freely between the epidermis cells. 

It was said (Netolitzky, p. .38) that the disappearance of the cuticle lining the 
inner walls of the micropyle seems to be the rule. But in some species we saw it 
either well presei-ved (PL xi, fig. la, 2a. 46) or in fragments (PL xi, fig. 4a). 

Relative Position of Chalaza, Hjlum, Michopyle .\ni) Hypocoxyl. 

In the hemitropous Eudesmiae the arrangement of these parts on the ventral side 
of the seed is illustrated by Figure 39. As can be seen, only in E. haileyana (PL xii, 
fig. 1; Text-fig. 39a) are they lying on the median line and symmetrical with it. In 
the other species they show considerable dislocations. Often the chalaza is turned 
away from the median and the hilum placed obliquely to it. As already mentioned, the 
position of the hilum determines always that of the micropyle: it lies on the prolonga- 
tion of the longitudinal hilum axis towards the root pole.* The position of the hypocotyl 
with the apical root meristem indicates — and germination tests confirm — that it does 
not emerge through the micropyle. 

In the anatropous Eudesmiae the disposition of those parts follows the Renantherae 
pattern (see Part I, Fig. 12); the chalaza lies on the raphe side just below the top. 
Hilum, micropyle and hypocotyl tip are at the basal end of the seed. 

NucELLUs, Endospekm and Embryo. 

Remnants of the nucellus tissue, the cells empty and obliterated, are always present 
in variable amounts and extent and line the inner face of the testa. The endosperm 
remnants enveloping the embryo are better preserved. Their thick-walled cells are still 
rich in proteins. 

Nucellus or endosperm (or both) project into the embryo folds between cotyledons 
and hypocotyl, thus showing ridges on the inner face. They may even penetrate into 
tliG wings (Fig. 28). The greatest amount is always found beneath the chalaza. 

Generally, in the anatropous Eudesmiae the nucellus lines the whole inner face 
of the testa while the endosperm is restricted to the chalazal and micropylar region. 
In the hemitropous Eudesmiae the inverse relationship exists: the endosperm can be 
isolated as an unbroken layer enveloping the whole embryo whilst the nucellus is 
confined chiefly to the ventral side of the seed. 

* This direction is anatomically well marked by elongated' rectangular epidermis ceU.s 
leaoing from the base of the hilum to the micropyle (Text-fig. 29, 30; PI. xi, fig. 5a; PI. xii, 
fig. 1). Only a histogenetic investigation would reveal what significance has to be attributed 
to this constant feature. 



BY E. GAUBA AND L. J). PKYOE. 289 

The embryo, covered by a semi-solid cuticle, consists of the two cotyledons folded in 
a complicated way and bent down along the hypocotyl. Root and shoot are still at a 
primordial stage, that is, as apical meristems occupying the two poles of the hypocotyl 
Avhich alone represents the embryo axis. 

The reserve materials stored in the embryo are oil droplets and aleurone grains, 
the latter enclosing cluster crystals of calcium oxalate and globoids. 

To exhibit the oxalate crystals it is advisable to remove successively the oil. proteins 
and globoids, whereupon they are clearly seen (up to 8 or more per cell). After treat- 
ment with HCl they disappear by formation of soluble CaCL, whereas cautious addition 
of sulphuric acid converts them on the spot into clusters of needle-shaped crystals of 
calcium sulphate. 

The distribution of the cluster crystals of calcium oxalate in the embryo cotyledons 
is also worthy of note. In E. erythrocorys, E. tetrac/ona, E. endesnioides, E. ehhanoensis 
and E. odontocarpa they occur in all parts of the mesophyll, whereas in E. haileyana 
and E. similis the palisade tissue is devoid of them, though in the latter clusters in a 
few palisades can occasionally be observed. In E. curtisii and E. tenuipes (as well as 
in Tristania conferta) they seem to be missing altogether. 

In Angophoras (A. intermedia, A. costata, A. cordifolia) they are confined to the 
upper epidermis, each cell containing one large druse or having instead a single crystal 
(tetragonal bipyramid?). Such crystals can be observed occasionally also among cluster 
crystals of E. baileyana, but here they are rare. 

The globoids are generally of globular shape, but irregular form: even vermicular 
(E. erythrocorys) can be encountered frequently. 

Is THE Series Euuesmiae a Natuual Grot:p? 

It is necessary to consider separately the Subseries with anatropous and hemi- 
tropous seeds in the light of the above results. 

The two anatropous species, E. tenuipes and E. curtisii, agree completely in all 
anatomical details with each other, in spite of the quite different shape of their seeds. 
In E. curtisii they are very slender, "somewhat similar to those of Tristania conferta 
R.Br., and quite distinct from any other species of Eucalyptus". In E. tenuipes they 
are "D-shaped to oblique pyramidal" (Blakely, 1930). Thus, they are "leptosperm" in 
E. ciirtisii only, but of a "renantherous" type in E. tenuipes. Nevertheless they are 
undoubtedly close relatives, but the designation "Leptospermae" for this Subseries is 
inappropriate as it covers only one of the two species. 

Amohg the hemitropous species the members of the Subseries Holocalyces: E. 
baileyana, E. similis and E. lirata* seem also to form a natural group. Blake (7oc. cit., 
p. 335) mentioned that similarities of some vegetative and floral characters "suggest 
that E. similis, E. baileyana (and presumably E. lirata), E. phoenicea and E. miniatu 
are better grouped together". We shall comment on this proposal after having investi- 
gated the Series Miniatae. 

In the Subseries Efasciculares the seeds of its two species, E. odontocarpa and 
E. tetrodonta, have little in common both anatomically and morphologically, and to 
unite them in a group on seed characters alone appears impossible. E. odontocarpa 
has thin, flat, somewhat concave-convex seeds, carinate on the back, and with a circum- 
ferential wing which, though tender and narrow, has nevertheless a very complex 
structure (Fig. 28). The surface is wrinkled due to bulging mucilaginous epidermis 
cells, and the hiluni sclerified (PI. xii, flg. 1). In E. tetrodonta the seeds are angular- 
orbicular in outline and are depressed and rather thick, costate with the ribs built up 
by macrosclereids, the surface often warty due to groups of lignified cells perpendicu- 
larly elongated to it. There is no trace of sclerification in the hilum. Among the 



* Of B. lirata we had two seeds only (from the type specimen) for examination .so thy.t 
not all structural details could be investigated. 



290 SEED COAT ANATOMY AND TAXONOMY IN EUCALYPTUS. II, 

heinitropous Eudesmiae E. odontocarpa is the only species with a straight, palmately 
branched chalaza bundle (PI. xii, fig. 1), whereas in E. tetrodonta the branches have 
an arctuous course (Fig. 37). On the other hand, E. tetrodonta is the only species in 
the Eudesmiae without any suberization in the chalaza region. Miiller (Eucalypto- 
graphia, Dec. I) reports that "E tetrodonta has no immediate close affinity to any of 
its congeners, except to E. odontocarpa . . .", and Blakely (Key, p. 70) says that "the 
two species are thrown together by the morphosis of the buds and fruits, but differing 
widely in habit". Blake {loc. cit., p. 335), commenting on Blakely's Eudesmiae, is 
cautious when suggesting that this Series "should perhaps be limited to E. tetragona, 
E. odontocarpa, E. erythroeorys, E. eudesmoides. and perhaps E. tetrodonta". 

A still greater diversity is encountered in the Fasciculares, comprising E. erythro- 
eorys, E. tetragona, E. eudesmoides and E. ehhanoensis. E. erythroeorys has solid ribs 
formed by strongly elongated, thick walled, lignifled epidermis cells, unique among the 
Eudesmiae (Fig. 22, 26). E. tet7-agona and E. eudesmoides have winged seeds, those of 
E. ehhanoensis are without any testa expansion. Thus, purely externally they al! 
display a very different aspect. Among the Eudesmiae E. erythroeorys is the only 
species with the rim of macrosclereids surrounding the hilum (Fig. 31) and with the 
most complicated reticular venation which in E. tetragona is very poor and simple 
(PI. xiii, fig. 2). Unique among all Eucalypts so far examined, and to our knowledge 
perhaps the Angiosperms hitherto investigated, is the suberization of the hilum in 
E. ehhanoensis (PI. xi, fig. 5). The micropyle structure of E. erythroeorys and E. 
tetragona is also unusual. The crystal epithelium is elongated into a tube or horn- 
like protuberance, sometimes extended up to the outer epidermis (PI. xi, fig. la), or 
even terminating freely between the epidermis cells (PI. xi, fig. 16). 

We have stressed here only some of the more striking differences in morphological 
and anatomical seed characters of species allotted by Blakely to four Subseries and 
these united into the Series Eudesmiae. In the light of the above evidence the question 
arises as to which facts induced Blakely to make this classification. He considers 
(Key, p. 68) the Eudesmiae as "closely allied to Angophora particularly in the opposite 
character of the leaves, toothed calyx, texture of some of the fruits, and in the 
morphology of the seeds". , 

So far as the phyllotaxis and the calyx teeth are concerned, none of these characters 
is constant throughout the whole Series. 

E. ehhanoensis, E. similis and E. haileyana are quoted as having alternate leaves. 
In E. tetragona, E. tetrodonta, E. lirata, E. tenuipes and E. curtisii opposite and 
alternate leaves occur. 

With regard to the calyx teeth Blakely's designation of the Subseries "Holocalyces" 
expresses their lack in this group. 

In view of the great variety of seed types in the Eudesmiae it is difficult to under- 
stand Blakely's hint at the morphological similarity of Angophora and Eucalyptus seed. 
Taking for instance the very different seeds of E. erythroeorys, E. tetragona, E. 
ehhanoensis, E. haileyana, E. curtisii or E. tenuipes; none resembles at all seeds of 
Angophora intermedia, A. costata or A. eordifolia. A cursory glance at the anatomical 
structure of these Angophoras reveals a type of epidermis not seen in any Eudesmiae 
(thin-walled palisade-like cells rich in tannins), the inner integument completelj^ or 
partially resorbed and not suberized, the oxalate druses of the cotyledons confined tc 
the upper epidermis, etc. 

Surprisingly, a comparison of Eucalyptus curtisii with Tristania con'ferta (both 
with anatropous seeds) reveals a perfect uniformity not only in all morphological 
details — as already pointed out by Blakely (Key, p. 72) — but also in the anatomical 
testa structure. 

At this stage of our investigations further speculation about the affinity of species 
is not justified. 



BY E. GAUISA AND L. D. PKYOli. 291 

References. 

(Additional to those cited in Part I, Proc. Linn. Soc. N.S.W., 195S, S3, Part 1, p. 31.) 
Baker, R. T., and Smith, H. G., 1920. — A Research on the Eucalypts (2nd edition). 
Blakhlt W. F., and White, C. T., 1930. — Two intere.sting Queensland Eucalypts. Proc. Roy. 

Soc. Queensl., 42, No. 9. 
HiNTZiNGER, A., 1927. — ^tjber die Ablosung der Samen von der Placenta etc., Sitsh. Akad. Wiss. 

Wien. Mathem.-nat. Kl., Abt. I: 136. 
KtJSTER, E., 1903. — Patholoyische P fianzenanatomie (1st edition). 
PURI, v., 1951. — The Role of Floral Anatomy in the Solution of Morphological Problems. Bot. 

Rev., XVII, 7. 

EXPLANATION OF PLATES XI-XIII. 

Plate xi. 

(Cork tissue and cuticles brown.) 

1-4: Micropylar structures. \a-c, E. e.rythrcorys : la, longit. sect. (semi-diaf;rammatic), 

45x ; lb, "crystal cone" protruding- between epidermis cells (semi-diagrammatic), 45x; Ic, 

endostome from la, 125x. 2a-Z), E. ehhanoe^isis, 330x : 2a, micropylar aperture in the outer 

epidermis; 2b, endostome (open). Za-h, E. tenuipes, surface view, 250x : 3a, endostome 

(closed) ; 3b, inner cuticle beneath the endostome. 4o-b, E. eudesvioides, surface view, 250x : 

4a, micropylar slit in the outer integument; 4b, micropylar slit in the crystal epithelium. 5a-b, 

E. ehbanoensis, hilurn, 160x: aa, surface view; 5b, trans, sect. 

Plate xii. 
Vascularization patterns (photomicrographs). 
1 : E. odontocarija, ca. lOOx. ch, clialaza ; U, hilum ; ni, micropyle. 2 : E. eudesnioides , ca. 
lOOx. 

Plate xiii. 
Vascularization patterns (photomicrographs). 
1: E. baileyana, ca. lOOx. 2: E. tetrafjoyia. ca. 90y. 



292 



EXPERIMENTAL CROSSING OF AEDES (8TEG0MYIA) AEGYPTI LINNAEUS AND 
AKDES iSTEGOMYIA) ALB0PICTV8 SKUSE (DIPTERA, CULICIDAE). 
By A. R. WooDiiiLL, Department of Zoology, University of Sydney, Sydney. 

(Plate xiv.) 

[Read 26th August, 1959.] 



Synopsis. 
Experimental crossings were made between various strains of A. aegypti and A. alboxiictus. 
Out of a total of i9,C49 eggs deposited all were sterile, with the exception of one egg which 
gave rise to a male adult which showed thoracic markings combining the characters of the 
parent species. 



IxTRODUCTIOjV. 

This cross was first attempted by Toumanoff in 1937 and since then by variouL^ 
workers; the results of these experiments have been fully discussed by Kitzmiller 
(1953) and by Mattingly (1956) and a complete list of references will be found in their 
papers. The results obtained by various authors may be briefly summarized as follows: 
in all cases where fertile progeny was obtained the reciprocal cross was sterile, while 
the Fi progeny and succeeding generations all completely resembled the female parenr 
with the exception of one female in Toumanoff's and one male from a back cross in 
Bonnet's experiments which resembled the male parent. In some crosses the fertile 
progeny were produced by using A. aegypti as the female parent and in others by using 
-4. anopictus as the female. Other workers again found complete sterility in both 
reciprocal crosses. In view of these contradictory results the author carried out a 
series of experiments during the period 1955-1959, the results of which are presented 
in this paper. 

Methods. 

Standard laboratory colonies of various strains of A. aegypti and A. alhopictus 
were maintained and eggs from these were used to obtain adults for the crossing 
experiments. Single pupae were isolated in tubes and the resulting adults were 
checked twice for species and sex before being liberated in the cages. Mass matings 
were made, with the numbers of males and females per cage varying from approximately 
100 to 250. In all experiments examinations were made of spermathecae of small 
samples of females for the presence of living spermatozoa. Eggs from the matings 
were deposited on rough textured white filter papers placed in small dishes of water, 
and these were kept wet for 48 hours to allow the embryos to develop and were then 
dried out slowly and kept at 80°F. and 90% relative humidity for 7 days before being 
counted and immersed in water; they were then kept in water to observe hatching for 
at least 10 days before being discarded. Small quantities of dried yeast and ground 
rat biscuits were added to the water, as it has been suggested that some organic 
pollution stimulates hatching, although this has not been the author's experience. The 
adults were kept in nylon marquisette cages 12" x 10" x 10", were given raisins, sugar 
solution and fruit, and offered blood feeds three times weekly. The breeding work was 
carried out in a warm room at 80°F. and 75% to 80% relative humidity with natural 
daylight, including some hours of direct sunlight. 

Eggs from the standard colonies treated as described above were 100% fertile. 
After the eggs from any particular culture or cross were handled, all glassware, 
instruments and bench surfaces were sterilized by heat treatment and the hands of 
the author thoroughly washed in very hot water. All cages were sterilized by dry 
heat before being used. These methods were also applied at all times to the handling 
of standard colonies. 

The work was carried out entirely by the author with the exception of counting: 
the eggs, which was carried out by an assistant who maintained the same precautions. 

Proceedings of the Liis'nean Society of New South Walks, 1959, Vol. Ixxxiv, Part 2. 



BY A. E. WOOD HILL. 



293 



Results. 

Dui'ing the period 1955-1957 crosses were attempted between A. aegypti from 
Mornington Island, Queensland (Q.), and two sti-ains of A. alhopictus, one from the 
Philippines (P.) and the other from Singapore (S.). Specimens of the strain of 
A. aegypti used were subm.itted to P. F. Mattingly, who stated that they were mainly 
var. queenslandensis with a few dark enough to be considered as the type form (see 
Mattingly, 1957). 

The details of the experiments are shown in Table 1. 

Table 1. 



Experi- 
ment 
Number. 


Date. 


Number of Mosquitoes, 
Sex and Species. 


Number 

of Blood 

Feeds. 


Number 

of Eggs 

Deposited. 


Number 
of Eggs 
Hatched. 


$$ Examined for 
Living Sperms. 


Number 
Positive. 


Number 
Negative. 


1 


June, 
1955 


254$ alb. (P) X 245cJ aeg. (Q). 
250? aeg. (Q) x 260(? alb. (P). 


370 

660 


177 
6,538 






3 

12 


22 
13 


2 


June, 
1956 


122$ alb. (S)xl33(?aeg. (Q). 
100$ aeg. (Q) x 105c? alb. (S). 


210 
220 


1,676 
349 


1 



4 

1 


6 
9 


3 


July, 
1957 


182$ alb. (P) X 241cJ aeg. (Q). 
186$ aeg. (Q) x 246(? alb. (P). 


285 
560 


1,565 
1,089 






1 



9 
10 


i 


Sept., 
1957 


180$ alb. (S)x216^ aeg. (Q). 
180$ aeg. (Q) x 243(J alb. (S). 


305 
435 


837 
8,757 






1 



19 
20 




20,988 





Copulation was observed frequently in all crosses and in most cases living 
spermatozoa were observed in at least some of the females. The females fed freely, 
but the number of eggs deposited did not always correspond with the number of blood 
feeds taken. The number of females examined for living spermatozoa was too small 
to allow of any correlation between these figures and the number of eggs deposited. 
It will be seen that out of a total of 20,988 eggs, only one was fertile. This was bred 
through to the adult stage and resulted in a male specimen which was completely 
intermediate in dorsal thoracic markings between A. aegypti and A. aWopicUis (see 
Plate xiv). The genitalia of this specimen were submitted to P. F. Mattingly, who 
considers that they resemble A. alhopictus with some slight modification in the direction 
of A. aegypti. 

The most recent report of successful crosses between A. aegypti and A. alhopictus 
is that by Bonnet in Hawaii in 1950. It was thought that the failure to repeat the 
results of previous experiments may have been due to the presence of different strains 
of the species used. Accordingly eggs of both species were obtained from Hawaii (H.), 
and the results of a series of crosses made with these during 1958-1959 are shown in 
Table 2. 



Table 2. 



Experi- 


Date. 


Number of Mosquitoes, 
Sex and Species. 


Number 
of Blood 

Feeds. 


Number 

of Eggs 

Deposited. 


Number 
of Eggs 
Hatched. 


$$ Examined for 
Living Sperms. 


Number. 


Number 
Positive. 


Number 

Negative. 


1 


Oct., 
1958 


106$ alb. (H) X 127cJ aeg. (H). 
128$ aeg. (H)xl82cJ alb. (H). 


172 
432 


3,169 
13,326 






5 
3 


5 

7 


2 


Feb., 
1959 


143$ alb. (H) X 188(? aeg. (H). 
140$ aeg. (H) X 194cJ alb. (H). 


151 
495 


3,227 
8,939 






12 
6 


s 

14 




28,661 





294 EXPERIMENTAL CROSSING OF AEDES AEGYPTI AND A. AEBOPICTXJS, 

As before, frequent copulation was observed in all crosses and at least some of 
the females showed living spermatozoa in the spermathecae. In this series A. aegypti 
fed much more freely than A. alhopictus. and this showed a distinct correlation with 
the number of eggs deposited. It will be seen that of a total of 28,661 eggs deposited 
all were completely sterile. 

DlSCVSSION. 

The experiments outlined above were carried out on a much larger scale than 
those made by previous workers and all possible precautions were taken to avoid 
contamination of cultures by stray eggs. Two different strains of A. aegypti were 
used and three different strains of A. alhopictus, though not in all the possible com- 
binations. It is surprising therefore that the results of previous workers could not be 
confirmed. This leads to consideration of the possibility that previous results may 
have been due to the. contamination of cultures by stray eggs accidentally introduced. 
As the eggs of these species remain viable in the dry state for several months there is 
an ever present danger, when various cultures are being bred in the same laboratory, 
that eggs may adhere to instruments, glassware or the hands of the operator or fall 
onto bench surfaces and be accidentally picked up again. The whole position is very 
puzzling and must remain open for the time being. It is hoped to obtain further strains 
of both species and to repeat the experiments with these. 

Acknowledgements. 
The author wishes to thank Mr. P. F. Mattingly for examining specimens of A. 
aegypti and the genitalia of the hybrid and for his continued interest in the work. 
Dr. L. E. Rozeboom, Dr. D. H. Colless, Mr. P. Y. Nakagawa and Dr. Elmo Hardy for 
forwarding eggs of various species, and Miss E. Hahn and Mr. R. Barnard for assistance. 
The latter was responsible for breeding through the single hybrid while the author was 
in hospital. 

References. 
KiTZMiLLER, J. B., 1953. — Mosquito Genetics and Cytogenetics. SeparaUi da Rev. Bras, de 

Malariol. e D. Ti-op., Vol. 5, No. 4: 285-359. 
Mattingly^ P. F., ]956. — Species Hybrids in Mosquitoes. Trans. Roy. Ent. Soc. London, lOS, 

Pt 2: 21-36. 
Mattingly. P. F., 1957. — Genetical Aspects of the .lerfes aegypti Problem. Ann. Trap. Med. and 

Parasit., 51, No. 4 : 392-408. 



EXPLANATION OF PLATK XIV. 
Adult male showing- combination of characters of A. aegypti and ^1. alhopictus. 




Piioc. Linn. Soc. N.S.W., 1959. 



Plate v. 




1, 2, Ctenomorphodes tessulatu. 3, M ijriiieconiiniesis sp. 



Pkoc. Linn. Soc. N.S.W., 1959. 



Plate vi. 










Pliasmaticl eggs. 



Proc. Liiv>;. Soc. N.S.W., 1959. 



Plate vii. 










^fS 






I' 



la 



''\^ 








Xylehorus truncatus. 



Proc. Linn. Soc. N.S.W., 1959. 



Plate Viii. 





B. 









■ %. 



m 




. G. 




Pitoc. Ltnn. Soc. N S.W., 1959, 



Plate ix. 




Cyinatilesta spenyleri. A, B. Female ancT egg mass. C. Veligers, 11 days. 



Pkoc. Linn. Soc. N.S.W., 1959. 



Plate x. 




Pollen tetrads in Leschenaultia. 



Peoc. LiiXN. Soc. N.S.W., 1959. 



PLATL X3. 




Pkoc. Linn. Soc. N.S.W., 1959. 



Plate Xri. 



/'■ 



IJ 



>K 






r \ 






,'\ 



^'y 



■^3,'t}^. J^^ii^' I! 





%: 



.m 







I'lioc. Linn. Soc. N.S.W., 1959. 







" ^ 
\i.^^ 



Plate xiii 

3 



'%\ 









-\ "«« 



- i* r 



:^^ 






ch 



/' 









"5^' dp l'-^. 






fi^. 1 






X 



.1 ^.^ 



N 



fC- 



% 



y^. 









-^ 



















':^ 







Pkoc. Linn. Soc. N.S.W., 1959. 



Plate xiv. 




Adult male 



showing combination of characters of Aedes aeciypti and Ae. albopictus. 



Notice to AtrrHOBS. 

Authors are requested to assist in facilitating the Society's publishing work by 
observing the following recommendations when preparing papers for submission to the 
Council. Perusal of the Pboceedings will show the general style to be adopted. 

Manuscript. — Manuscripts should be double-space typed with a margin of at least one 
inch on the left-hand side, top and bottom of each page. The title should be spaced 
clearly above the text on the first page. The original should be submitted and a 
copy retained by the author for checking proofs. A Table of Contents, on a separate 
sheet, should accompany each manuscript; this is not necessarily for publication, 
but will serve to show the proper relation of the headings. A manuscript when 
submitted to the Council should be complete in every detail, both with regard to 
text, references and illustrations, for any extensive alterations or corrections made 
on the proofs, if allowed, will be at the author's expense. No words except generic 
and specific names, and those to be printed in italics, should be underlined. An 
abstract of the paper should accompany the manuscript. 

References. — References should be carefully checked by the author, who is alone respon- 
sible for their accuracy. They should be listed alphabetically at the end of the 
manuscript, and should be cited in the text by the author's name, e.g., Bullough 
(1939) or (Bullough, 1939); and in the list thus: 

Bullough, W. S., 1939. — A Study of the Reproductive Cycle of the Minnow in 
Relation to the Environment Proc. Zool. Soc. Lond., 109, A, Pt. 1:79-108. 

Abbreviations. — Standard abbreviations should be used in tabulations and after numerals 
in the text. The abbreviations of names of periodicals should conform to those in 
the World List of Scientific Periodicals. 

Tabulations. — Tables should be numbered consecutively and referred to specifically in 
the text by number. Each table, provided with a heading descriptive of the contents,, 
should be submitted on' a separate sheet. 

Illustrations. 

Plates. — The size should not exceed 7i x 5 in. except if the subject will bear reduction. 
A number of small photographs should be arranged to make one plate. Photographs 
should show good contrast and be printed on glossy paper. 

Line drawings will, as far as possible, be printed in the text. Drawings should be 
made on white board or stiff white paper with Indian ink. It is advisable to arrange 
a number of text-figures, all to bear the same reduction, on a single large sheet, which 
can be reduced to one-third, one-half or one-quarter the original size, giving a block 
8 X 5 in. or 4 X 5 in. without loss of essential detail! Each separate figure should be 
clearly numbered and all lettering should be plain and large enough to be clearly 
readable when reduced. If co-ordinate paper is used for graphs it should be blue- 
lined. The explanation of Text-figures should be supplied on a separate sheet; but 
the explanation of Plates should be given at the end of the manuscript. Text-figures 
and Plates should be numbered consecutively and referred to in the text by number. 
Indicate in the manuscript where each Text-figure or group of figures is to be inserted. 

Proofs should be corrected and returned as soon as possible together with an order for 
any reprints required in addition to the 50 supplied gratis by the Society. 

Synopsis. — A synopsis of the contents of each paper must accompany the manuscript, 
to appear below the title. Such a synopsis is to be regarded as an essential part 
of a paper and is to be supplied by the author. It does not take the place of the 
usual Abstract which is used for inclusion in the Society's Monthly Abstract series. 
Each synopsis should state clearly the contents of the paper and outline the main 
conclusions or findings. In taxonomic papers reference should be made to new 
genera and species described if these appear in the paper. 



PROCEEDINGS, LXXXIV, PART 2, 1959 



CONTENTS. 

Pages 

Observations on the Ecology of the Phasmatid Ctenomorphodes tessulata 

(Gray). By P. Hadlington and F. Hoschke. (Plates v-vi; two Text-figures.) 146-159 

An Annotated Catalogue of Described Australian Tabaninae (Diptera, 

Tabanidae). By I. M. Mackerras. (Nine Text-flgures.) 160-185 

Observations on Some Australian Forest Insects. 4. Xylehorus truncatus 
Erichson, 1842 (Coleoptera: Scolytidae) associated with Dying Eucalyptus 
saligna Smith (Sydney Blue-gum). By K. M. Moore. (Plate vii; three 
Text-figures.) 186-193 

The Genus Conostylis R.Br. I. Leaf Anatomy. By J. W. Green. (Twenty-five 

Text-figures.) 194-206 

Somatic Hybridization between Puccinia graminis var. tritici and Puccinia 

graminis var. secalis. By I. A. Watson and N. H. Luig 207-208 

Diptera of Katoomba. Part 3. Stratiomyiidae and Tachinidae. By G. H. Hardy 209-217 

Some East Australian Sea-grass Communities. By E. J. Ferguson Wood. (Two 

Text-figures.) 218-226 

The Effects of Inorganic Salts on Dividing Cells. By Mary M. Hindmarsh. 

(Plate viii; two Text-figures.) 227-231 

The Reproduction and Early Life History of the Gastropod Cymatilesta 
spengleri (Perry) (Fam. Cymatidae). By D. T. Anderson. {Communi- 
cated by Miss I. Bennett.) (Plate ix; seventeen Text-flgures.) 232-237 

A New Mite Parasite (Harpy rhynchus) from the Roselle Parakeet (Trombidi- 
formes, Acari). By R. F. Lawrence. [Communicated by Dr. A. R. 
Woodhill.) (Two Text-figures.) 238-241 

Floral Structure and Anatomy in the Family Goodeniaceae Dumort. By R. C. 

Carolin. (Forty-five Text-figures.) 242-255 

Iron Deficiency in Eucalyptus dives Schauer. By W. D. Andrew and D. J. 

David. (Communicated by Professor L. D. Pryor.) 256-258 

Pollen Tetrad Segregation in Astroloma pinifolium and in Acrotriche fasciculi- 
flora. By S. Smith-White. (Four Text-flgures.) 259-270 

Pollen Tetrad Patterns in Leschenaultia. By P. G. Martin and W. J. Peacock. 

(Plate X.) 271-277 

Seed Coat Anatomy and Taxonomy in Eucalyptus. II. By E. Gauba and L. D. 

Pryor. (Plates xi-xiii; Text-figures 20-39 I 278-291 

Experimental Crossing of Aedes (Stegomyia. gypti Linnaeus and Aedes 
(Stegomyia) albopictus Skuse (Diptera, Cai oidae). By A. R. Woodhill. 
(Plate xiv.) 292-294 



,^/^^./^' 



(Issued, 22nd February, 1960.) 











DC=0 


Vol. LXXXIV. ^*'- 391- 
Part 3.. 

THE 

PROCFEDINGS 

OF THB 

1 ,I^^En^ SOC E lY 

or 

New South Wales 

FOR THB YEAR 
1959 










Part 3 (pp. 295-432). 










COIVTAIXING PAPERS READ IIV SEPTEMBER-NOVEMBER. 










With nine plates. 
[Plates xv-xxiii.] 










Stdn et: 










PRINTED AND PUBLISHED FOB THE SOCIETY BY 








s 


AUSTRALASIAN MEDICAL PUBLISHING CO. LTD., 

Seamer and Arundel Streets, Glebe, Sydney, 

and 

SOLD BY THE SOCIETY, 

Science House, 157 Gloucester Street, Sydney. 


I 






Registered at the General Post Office, Sydney, for transmission 
by post as a periodical. 






Agent in Europe: 
David Nutt, 41 Colebrooke Row, London, N.l. 





The Linnean Society of New South Wales 



LIST OF OFFICERS AND COUNCIL, 1959-60. 



President; 

T. G. Vallance, B.Sc, Ph.D. 

Vice-Presidents s 

S. J. Copland, M.Sc. F. V. Mercer, B.Sc, Ph.D. 

Lilian Fraser, D.Sc. S. Smith-White, D.Sc.Agr. 

Hon. Treasnrer: 

A. B. Walkom, D.Sc. 

Hon.. Secretaries: 

W. R. Browne, D.Sc; A. B. Walkom, D.Sc. 

Council: 

R. H. Anderson, B.ScAgr. S. Smith-White, D.Sc.Agr. 

W. R. Browne, D.Sc. E. Le G. Troughton, C.M.Z S., F.R.Z.S. 

A. N. Colefax. B.Sc. T. G. Vallance, B.Sc, Ph.D. 

S. J. Copland, M.Sc J. M. Vincent, D.Sc.Agr., Dip.Bact. 

Professor R. L. Crocker, D.Sc. A. B. Walkom, D.Sc. 

J. W. Evans, M.A., D.Sc, Sc.D. H. S. H. Wardlaw, D.Sc, F.R.A.C.I. 

Lilian Fraser, D.Sc Professor W. L. Waterhouse, C.M.G., 
F. V. Mercer, B.Sc, Ph.D. M.C., D.Sc.Agr., D.I.C. 

I. V. Newman, M.Sc, Ph.D. A. R. Woodhill, D.Sc.Agr. 
Professor B. .1. F. Ralph, B.Sc, Ph.D., 
A.A.C.I. 

Auditor: 

S. J. Rayment, F.C.A. (Aust.). 



NOTICE 

Sets of the Feoceedings of the Linnean Society of New South Wales (except 
First Series, Volumes I-VI, 1875-1881), and separate Parts, may be purchased from the 
Society, Science House, 157 Gloucester Street. Sydney. Separate Parts may be obtained 
also from David Nutt, 41 Colebrooke Row, London, N.l. The stock of First Series, 
Volumes I-VI, is strictly limited and some Parts of these Volumes are out of print. 

From 1st March, 1954, the price of each complete volume will be £3 10s., plus postage. 
(All previous price lists are cancelled.) 

Index to Volumes I-L of the Proceedings [Issued 15th February, 1929]. Pp. 108. 
Price 5s. 

The Macleat Memorial Volume [Issued 13th October, 1893]. Royal 4to, ii to 308 
pages, with portrait, and forty-two plates. Price £1 Is. 

Descriptive Catalogue of Australian Fishes. By William Macleay, F.L.S. [1881]. 
A few copies only. Two volumes and supplement. Price £2 2s. 

The Transactions of the Entomological Society of New South Wales, 2 vols., 8vo. 
[Vol. I (complete in five parts, 1863-66), price 70s. net. Parts 2-5 IDs. each; Vol. II 
(complete in five parts, 1869-73), price 30s. net, or single Parts 7s. 6d. each.] 



SUBSCRIPTIOjV: £S 10s. per annum: po.stage 4s. 



295 



RANUNCULUS LAPPACEU8 AND ALLIED SPECIES OF THE AUSTRALIAN 

MAINLAND. 

I. Taxonomy. 

By Barbara G. Beiggs, Department of Botany, University of Sydney. 

(Plate XV ; Ninety-one Text-figures.) 

[Read 30th September, 1959.] 




Synoiisis. 
The species of the R. lappaceus group which occur on the Australian mainland are 
considered in the light of morphological and experimental data. R. colonorwm Bndl., 
previously reduced